[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <CA+h21hp8ANbZw63UAy5cDx389j6pSOHUG7t1mR+rbQcrFLROyw@mail.gmail.com>
Date: Wed, 10 Jun 2020 17:37:22 +0300
From: Vladimir Oltean <olteanv@...il.com>
To: Kurt Kanzenbach <kurt.kanzenbach@...utronix.de>
Cc: linux-rt-users@...r.kernel.org,
lkml <linux-kernel@...r.kernel.org>, rcu@...r.kernel.org,
Colin King <colin.king@...onical.com>
Subject: Re: stress-ng --hrtimers hangs system
Hi Kurt,
On Wed, 10 Jun 2020 at 08:58, Kurt Kanzenbach
<kurt.kanzenbach@...utronix.de> wrote:
>
> Hi Vladimir,
>
> On Tue Jun 09 2020, Vladimir Oltean wrote:
> > Just out of curiosity, what and how many CPU cores does your ARM64 box
> > have, and what frequency are you running them at?
> > Mine is a dual-core A72 machine running at 1500 MHz.
>
> That particular machine has a dual core Cortex A53 running at 1GHz.
>
> Thanks,
> Kurt
Ok, I just noticed that you said "v5.6*-rt*".
So I re-tested that 5.4 downstream kernel again that hanged before,
just applied the -rt1 patchset on top. And it's actually working fine
now, no hangs, no RCU stalls, no need to reduce sched_rt_runtime_us.
So maybe it's something related to the config?
Thanks,
-Vladimir
Powered by blists - more mailing lists