Re: 4.14-rt timer issues using PREEMPT_RT_FULL=y and NO_HZ_FULL_ALL=y

[bert schulze <spambemyguest@googlemail.com>]

Sebastian Andrzej Siewior (Thu, Dec 14, 2017 at 05:25:03PM +0100):
> On 2017-12-13 20:14:54 [+0100], bert schulze wrote:
> > Yes as timer migration kicked in on boot causing the interrupt storm.
> > Disabling it all together there are no more hangs on boot.
> 
> and with the patch I Cced you one we have fix for that but there is
> more…
> 
> > This is sleep from coreutils, I'm not setting up anything here.
> > top, cyclictest whatever using nanosleep is broken.
> 
> that sleep should be also affected.
> 
> > Tried that as mentioned, no interrupt storm and nothing fails to expire.
> > (Add mainline 4.14 upto 4.14.5 to the list of working versions).
> …
> 
> > Do you need anything else that would be helpful?
> 
> bisect with RT is an exercise of its own.
> It looks that if you set rcu_normal_after_boot to 0 in
> kernel/rcu/update.c then other problem should go away and everything is
> "normal" now, right?

System boots reliable with mentioned patch applied, and nanosleep is
working as well! I do not have to set rcu_normal_after_boot to 0 or what
is its purpose?

> Now I need to figure why that is broken…

Guess you already did :)

root@deb9:~# bash test 
# uname -a
Linux deb9.virt 4.14.3-rt5-rt+ #5 SMP PREEMPT RT Thu Dec 14 18:39:27 CET 2017 x86_64 GNU/Linux
# cat /proc/sys/kernel/timer_migration
1
# dmesg | grep hrtimer
# date
Thu Dec 14 18:41:21 CET 2017
# grep LOC /proc/interrupts
LOC:       8002        208        138         68   Local timer interrupts
# for cpu in {0..3} ;do time taskset -ac  sleep 0.1 ;done
real    0m0.103s
real    0m0.105s
real    0m0.105s
real    0m0.105s
# date
Thu Dec 14 18:41:22 CET 2017
# grep LOC /proc/interrupts
LOC:       8437        209        139         70   Local timer interrupts

Tested on bare hardware and QEMU, top and cyclictest are working as well
no more interrupt storms can be observed!

> 
> Sebastian
Thank you!
Bert