From: Jonathan Cameron <Jonathan.Cameron@huawei.com>
To: David Miller <davem@davemloft.net>
Cc: <paulmck@linux.vnet.ibm.com>, <dzickus@redhat.com>,
<sfr@canb.auug.org.au>, <linuxarm@huawei.com>,
<npiggin@gmail.com>, <abdhalee@linux.vnet.ibm.com>,
<sparclinux@vger.kernel.org>, <akpm@linux-foundation.org>,
<linuxppc-dev@lists.ozlabs.org>,
<linux-arm-kernel@lists.infradead.org>
Subject: Re: RCU lockup issues when CONFIG_SOFTLOCKUP_DETECTOR=n - any one else seeing this?
Date: Wed, 26 Jul 2017 18:13:12 +0100 [thread overview]
Message-ID: <20170726181312.0000040d@huawei.com> (raw)
In-Reply-To: <20170726.095432.169004918437663011.davem@davemloft.net>
On Wed, 26 Jul 2017 09:54:32 -0700
David Miller <davem@davemloft.net> wrote:
> From: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
> Date: Wed, 26 Jul 2017 08:49:00 -0700
>
> > On Wed, Jul 26, 2017 at 04:33:40PM +0100, Jonathan Cameron wrote:
> >> Didn't leave it long enough. Still bad on 4.10-rc7 just took over
> >> an hour to occur.
> >
> > And it is quite possible that SOFTLOCKUP_DETECTOR=y and HZ_PERIODIC=y
> > are just greatly reducing the probability of the problem rather than
> > completely preventing it.
> >
> > Still, hopefully useful information, thank you for the testing!
Not sure it actually gives us much information, but no issues yet
with a simple program running every cpu that wakes up every 3 seconds.
Will leave it running overnight and report back in the morning.
>
> I guess that invalidates my idea to test reverting recent changes to
> the tick-sched.c code... :-/
>
> In NO_HZ_IDLE mode, what is really supposed to happen on a completely
> idle system?
>
> All the cpus enter the idle loop, have no timers programmed, and they
> all just go to sleep until an external event happens.
>
> What ensures that grace periods get processed in this regime?
WARNING: multiple messages have this Message-ID (diff)
From: Jonathan Cameron <Jonathan.Cameron@huawei.com>
To: linux-arm-kernel@lists.infradead.org
Subject: Re: RCU lockup issues when CONFIG_SOFTLOCKUP_DETECTOR=n - any one else seeing this?
Date: Wed, 26 Jul 2017 17:13:12 +0000 [thread overview]
Message-ID: <20170726181312.0000040d@huawei.com> (raw)
In-Reply-To: <20170726.095432.169004918437663011.davem@davemloft.net>
On Wed, 26 Jul 2017 09:54:32 -0700
David Miller <davem@davemloft.net> wrote:
> From: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
> Date: Wed, 26 Jul 2017 08:49:00 -0700
>
> > On Wed, Jul 26, 2017 at 04:33:40PM +0100, Jonathan Cameron wrote:
> >> Didn't leave it long enough. Still bad on 4.10-rc7 just took over
> >> an hour to occur.
> >
> > And it is quite possible that SOFTLOCKUP_DETECTOR=y and HZ_PERIODIC=y
> > are just greatly reducing the probability of the problem rather than
> > completely preventing it.
> >
> > Still, hopefully useful information, thank you for the testing!
Not sure it actually gives us much information, but no issues yet
with a simple program running every cpu that wakes up every 3 seconds.
Will leave it running overnight and report back in the morning.
>
> I guess that invalidates my idea to test reverting recent changes to
> the tick-sched.c code... :-/
>
> In NO_HZ_IDLE mode, what is really supposed to happen on a completely
> idle system?
>
> All the cpus enter the idle loop, have no timers programmed, and they
> all just go to sleep until an external event happens.
>
> What ensures that grace periods get processed in this regime?
WARNING: multiple messages have this Message-ID (diff)
From: Jonathan.Cameron@huawei.com (Jonathan Cameron)
To: linux-arm-kernel@lists.infradead.org
Subject: RCU lockup issues when CONFIG_SOFTLOCKUP_DETECTOR=n - any one else seeing this?
Date: Wed, 26 Jul 2017 18:13:12 +0100 [thread overview]
Message-ID: <20170726181312.0000040d@huawei.com> (raw)
In-Reply-To: <20170726.095432.169004918437663011.davem@davemloft.net>
On Wed, 26 Jul 2017 09:54:32 -0700
David Miller <davem@davemloft.net> wrote:
> From: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
> Date: Wed, 26 Jul 2017 08:49:00 -0700
>
> > On Wed, Jul 26, 2017 at 04:33:40PM +0100, Jonathan Cameron wrote:
> >> Didn't leave it long enough. Still bad on 4.10-rc7 just took over
> >> an hour to occur.
> >
> > And it is quite possible that SOFTLOCKUP_DETECTOR=y and HZ_PERIODIC=y
> > are just greatly reducing the probability of the problem rather than
> > completely preventing it.
> >
> > Still, hopefully useful information, thank you for the testing!
Not sure it actually gives us much information, but no issues yet
with a simple program running every cpu that wakes up every 3 seconds.
Will leave it running overnight and report back in the morning.
>
> I guess that invalidates my idea to test reverting recent changes to
> the tick-sched.c code... :-/
>
> In NO_HZ_IDLE mode, what is really supposed to happen on a completely
> idle system?
>
> All the cpus enter the idle loop, have no timers programmed, and they
> all just go to sleep until an external event happens.
>
> What ensures that grace periods get processed in this regime?
next prev parent reply other threads:[~2017-07-26 17:13 UTC|newest]
Thread overview: 241+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-07-25 11:32 RCU lockup issues when CONFIG_SOFTLOCKUP_DETECTOR=n - any one else seeing this? Jonathan Cameron
2017-07-25 12:26 ` Nicholas Piggin
2017-07-25 12:26 ` Nicholas Piggin
2017-07-25 13:46 ` Paul E. McKenney
2017-07-25 13:46 ` Paul E. McKenney
2017-07-25 13:46 ` Paul E. McKenney
2017-07-25 14:42 ` Jonathan Cameron
2017-07-25 14:42 ` Jonathan Cameron
2017-07-25 14:42 ` Jonathan Cameron
2017-07-25 15:12 ` Paul E. McKenney
2017-07-25 15:12 ` Paul E. McKenney
2017-07-25 15:12 ` Paul E. McKenney
2017-07-25 16:52 ` Jonathan Cameron
2017-07-25 16:52 ` Jonathan Cameron
2017-07-25 16:52 ` Jonathan Cameron
2017-07-25 21:10 ` David Miller
2017-07-25 21:10 ` David Miller
2017-07-25 21:10 ` David Miller
2017-07-26 3:55 ` Paul E. McKenney
2017-07-26 3:55 ` Paul E. McKenney
2017-07-26 3:55 ` Paul E. McKenney
2017-07-26 4:02 ` David Miller
2017-07-26 4:02 ` David Miller
2017-07-26 4:02 ` David Miller
2017-07-26 4:12 ` Paul E. McKenney
2017-07-26 4:12 ` Paul E. McKenney
2017-07-26 4:12 ` Paul E. McKenney
2017-07-26 8:16 ` Jonathan Cameron
2017-07-26 8:16 ` Jonathan Cameron
2017-07-26 8:16 ` Jonathan Cameron
2017-07-26 9:32 ` Jonathan Cameron
2017-07-26 9:32 ` Jonathan Cameron
2017-07-26 9:32 ` Jonathan Cameron
2017-07-26 12:28 ` Jonathan Cameron
2017-07-26 12:28 ` Jonathan Cameron
2017-07-26 12:28 ` Jonathan Cameron
2017-07-26 12:49 ` Jonathan Cameron
2017-07-26 12:49 ` Jonathan Cameron
2017-07-26 12:49 ` Jonathan Cameron
2017-07-26 14:14 ` Paul E. McKenney
2017-07-26 14:14 ` Paul E. McKenney
2017-07-26 14:14 ` Paul E. McKenney
2017-07-26 14:23 ` Jonathan Cameron
2017-07-26 14:23 ` Jonathan Cameron
2017-07-26 14:23 ` Jonathan Cameron
2017-07-26 15:33 ` Jonathan Cameron
2017-07-26 15:33 ` Jonathan Cameron
2017-07-26 15:33 ` Jonathan Cameron
2017-07-26 15:49 ` Paul E. McKenney
2017-07-26 15:49 ` Paul E. McKenney
2017-07-26 15:49 ` Paul E. McKenney
2017-07-26 16:54 ` David Miller
2017-07-26 16:54 ` David Miller
2017-07-26 16:54 ` David Miller
2017-07-26 17:13 ` Jonathan Cameron [this message]
2017-07-26 17:13 ` Jonathan Cameron
2017-07-26 17:13 ` Jonathan Cameron
2017-07-27 7:41 ` Jonathan Cameron
2017-07-27 7:41 ` Jonathan Cameron
2017-07-27 7:41 ` Jonathan Cameron
2017-07-26 17:50 ` Paul E. McKenney
2017-07-26 17:50 ` Paul E. McKenney
2017-07-26 17:50 ` Paul E. McKenney
2017-07-26 22:36 ` Paul E. McKenney
2017-07-26 22:36 ` Paul E. McKenney
2017-07-26 22:36 ` Paul E. McKenney
2017-07-26 22:45 ` David Miller
2017-07-26 22:45 ` David Miller
2017-07-26 22:45 ` David Miller
2017-07-26 23:15 ` Paul E. McKenney
2017-07-26 23:15 ` Paul E. McKenney
2017-07-26 23:15 ` Paul E. McKenney
2017-07-26 23:22 ` David Miller
2017-07-26 23:22 ` David Miller
2017-07-26 23:22 ` David Miller
2017-07-27 1:42 ` Paul E. McKenney
2017-07-27 1:42 ` Paul E. McKenney
2017-07-27 1:42 ` Paul E. McKenney
2017-07-27 4:34 ` Nicholas Piggin
2017-07-27 4:34 ` Nicholas Piggin
2017-07-27 4:34 ` Nicholas Piggin
2017-07-27 12:49 ` Paul E. McKenney
2017-07-27 12:49 ` Paul E. McKenney
2017-07-27 12:49 ` Paul E. McKenney
2017-07-27 13:49 ` Jonathan Cameron
2017-07-27 13:49 ` Jonathan Cameron
2017-07-27 13:49 ` Jonathan Cameron
2017-07-27 16:39 ` Jonathan Cameron
2017-07-27 16:39 ` Jonathan Cameron
2017-07-27 16:39 ` Jonathan Cameron
2017-07-27 16:52 ` Paul E. McKenney
2017-07-27 16:52 ` Paul E. McKenney
2017-07-27 16:52 ` Paul E. McKenney
2017-07-28 7:44 ` Jonathan Cameron
2017-07-28 7:44 ` Jonathan Cameron
2017-07-28 7:44 ` Jonathan Cameron
2017-07-28 12:54 ` Boqun Feng
2017-07-28 12:54 ` Boqun Feng
2017-07-28 12:54 ` Boqun Feng
2017-07-28 13:13 ` Jonathan Cameron
2017-07-28 13:13 ` Jonathan Cameron
2017-07-28 13:13 ` Jonathan Cameron
2017-07-28 14:55 ` Paul E. McKenney
2017-07-28 14:55 ` Paul E. McKenney
2017-07-28 14:55 ` Paul E. McKenney
2017-07-28 18:41 ` Paul E. McKenney
2017-07-28 18:41 ` Paul E. McKenney
2017-07-28 18:41 ` Paul E. McKenney
2017-07-28 19:09 ` Paul E. McKenney
2017-07-28 19:09 ` Paul E. McKenney
2017-07-28 19:09 ` Paul E. McKenney
2017-07-30 13:37 ` Boqun Feng
2017-07-30 13:37 ` Boqun Feng
2017-07-30 13:37 ` Boqun Feng
2017-07-30 16:59 ` Paul E. McKenney
2017-07-30 16:59 ` Paul E. McKenney
2017-07-30 16:59 ` Paul E. McKenney
2017-07-29 1:20 ` Boqun Feng
2017-07-29 1:20 ` Boqun Feng
2017-07-29 1:20 ` Boqun Feng
2017-07-28 18:42 ` David Miller
2017-07-28 18:42 ` David Miller
2017-07-28 18:42 ` David Miller
2017-07-28 13:08 ` Jonathan Cameron
2017-07-28 13:08 ` Jonathan Cameron
2017-07-28 13:24 ` Jonathan Cameron
2017-07-28 13:24 ` Jonathan Cameron
2017-07-28 13:24 ` Jonathan Cameron
2017-07-28 16:55 ` Paul E. McKenney
2017-07-28 16:55 ` Paul E. McKenney
2017-07-28 17:27 ` Jonathan Cameron
2017-07-28 17:27 ` Jonathan Cameron
2017-07-28 17:27 ` Jonathan Cameron
2017-07-28 19:03 ` Paul E. McKenney
2017-07-28 19:03 ` Paul E. McKenney
2017-07-28 19:03 ` Paul E. McKenney
2017-07-31 11:08 ` Jonathan Cameron
2017-07-31 11:08 ` Jonathan Cameron
2017-07-31 11:08 ` Jonathan Cameron
2017-07-31 15:04 ` Paul E. McKenney
2017-07-31 15:04 ` Paul E. McKenney
2017-07-31 15:04 ` Paul E. McKenney
2017-07-31 15:27 ` Jonathan Cameron
2017-07-31 15:27 ` Jonathan Cameron
2017-07-31 15:27 ` Jonathan Cameron
2017-08-01 18:46 ` Paul E. McKenney
2017-08-01 18:46 ` Paul E. McKenney
2017-08-01 18:46 ` Paul E. McKenney
2017-08-02 16:25 ` Jonathan Cameron
2017-08-02 16:25 ` Jonathan Cameron
2017-08-02 16:25 ` Jonathan Cameron
2017-08-15 15:47 ` Paul E. McKenney
2017-08-15 15:47 ` Paul E. McKenney
2017-08-15 15:47 ` Paul E. McKenney
2017-08-16 1:24 ` Jonathan Cameron
2017-08-16 1:24 ` Jonathan Cameron
2017-08-16 1:24 ` Jonathan Cameron
2017-08-16 12:43 ` Michael Ellerman
2017-08-16 12:43 ` Michael Ellerman
2017-08-16 12:43 ` Michael Ellerman
2017-08-16 12:56 ` Paul E. McKenney
2017-08-16 12:56 ` Paul E. McKenney
2017-08-16 12:56 ` Paul E. McKenney
2017-08-16 15:31 ` Nicholas Piggin
2017-08-16 15:31 ` Nicholas Piggin
2017-08-16 15:31 ` Nicholas Piggin
2017-08-16 16:27 ` Paul E. McKenney
2017-08-16 16:27 ` Paul E. McKenney
2017-08-16 16:27 ` Paul E. McKenney
2017-08-17 13:55 ` Michael Ellerman
2017-08-17 13:55 ` Michael Ellerman
2017-08-17 13:55 ` Michael Ellerman
2017-08-20 4:45 ` Nicholas Piggin
2017-08-20 4:45 ` Nicholas Piggin
2017-08-20 4:45 ` Nicholas Piggin
2017-08-20 5:01 ` David Miller
2017-08-20 5:01 ` David Miller
2017-08-20 5:01 ` David Miller
2017-08-20 5:04 ` Paul E. McKenney
2017-08-20 5:04 ` Paul E. McKenney
2017-08-20 5:04 ` Paul E. McKenney
2017-08-20 13:00 ` Nicholas Piggin
2017-08-20 13:00 ` Nicholas Piggin
2017-08-20 13:00 ` Nicholas Piggin
2017-08-20 18:35 ` Paul E. McKenney
2017-08-20 18:35 ` Paul E. McKenney
2017-08-20 18:35 ` Paul E. McKenney
2017-08-20 21:14 ` Paul E. McKenney
2017-08-20 21:14 ` Paul E. McKenney
2017-08-20 21:14 ` Paul E. McKenney
2017-08-21 0:52 ` Nicholas Piggin
2017-08-21 0:52 ` Nicholas Piggin
2017-08-21 0:52 ` Nicholas Piggin
2017-08-21 6:06 ` Nicholas Piggin
2017-08-21 6:06 ` Nicholas Piggin
2017-08-21 6:06 ` Nicholas Piggin
2017-08-21 10:18 ` Jonathan Cameron
2017-08-21 10:18 ` Jonathan Cameron
2017-08-21 10:18 ` Jonathan Cameron
2017-08-21 14:19 ` Nicholas Piggin
2017-08-21 14:19 ` Nicholas Piggin
2017-08-21 14:19 ` Nicholas Piggin
2017-08-21 15:02 ` Jonathan Cameron
2017-08-21 15:02 ` Jonathan Cameron
2017-08-21 15:02 ` Jonathan Cameron
2017-08-21 20:55 ` David Miller
2017-08-21 20:55 ` David Miller
2017-08-21 20:55 ` David Miller
2017-08-22 7:49 ` Jonathan Cameron
2017-08-22 7:49 ` Jonathan Cameron
2017-08-22 7:49 ` Jonathan Cameron
2017-08-22 8:51 ` Abdul Haleem
2017-08-22 8:51 ` Abdul Haleem
2017-08-22 8:51 ` Abdul Haleem
2017-08-22 15:26 ` Paul E. McKenney
2017-08-22 15:26 ` Paul E. McKenney
2017-08-22 15:26 ` Paul E. McKenney
2017-09-06 12:28 ` Paul E. McKenney
2017-09-06 12:28 ` Paul E. McKenney
2017-09-06 12:28 ` Paul E. McKenney
2017-08-22 0:38 ` Paul E. McKenney
2017-08-22 0:38 ` Paul E. McKenney
2017-08-22 0:38 ` Paul E. McKenney
2017-07-31 11:09 ` Jonathan Cameron
2017-07-31 11:09 ` Jonathan Cameron
2017-07-31 11:09 ` Jonathan Cameron
2017-07-31 11:55 ` Jonathan Cameron
2017-07-31 11:55 ` Jonathan Cameron
2017-07-31 11:55 ` Jonathan Cameron
2017-08-01 10:53 ` Jonathan Cameron
2017-08-01 10:53 ` Jonathan Cameron
2017-08-01 10:53 ` Jonathan Cameron
2017-07-26 16:48 ` David Miller
2017-07-26 16:48 ` David Miller
2017-07-26 16:48 ` David Miller
2017-07-26 3:53 ` Paul E. McKenney
2017-07-26 3:53 ` Paul E. McKenney
2017-07-26 3:53 ` Paul E. McKenney
2017-07-26 7:51 ` Jonathan Cameron
2017-07-26 7:51 ` Jonathan Cameron
2017-07-26 7:51 ` Jonathan Cameron
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20170726181312.0000040d@huawei.com \
--to=jonathan.cameron@huawei.com \
--cc=abdhalee@linux.vnet.ibm.com \
--cc=akpm@linux-foundation.org \
--cc=davem@davemloft.net \
--cc=dzickus@redhat.com \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linuxarm@huawei.com \
--cc=linuxppc-dev@lists.ozlabs.org \
--cc=npiggin@gmail.com \
--cc=paulmck@linux.vnet.ibm.com \
--cc=sfr@canb.auug.org.au \
--cc=sparclinux@vger.kernel.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.