Mailinglist Archive: opensuse-bugs (4655 mails)

< Previous Next >
[Bug 1042933] kernel panic due to nmi caused by systemd-watchdog test
  • From: bugzilla_noreply@xxxxxxxxxx
  • Date: Wed, 21 Jun 2017 10:07:35 +0000
  • Message-id: <bug-1042933-21960-BlFV1MK48M@http.bugzilla.suse.com/>
http://bugzilla.suse.com/show_bug.cgi?id=1042933
http://bugzilla.suse.com/show_bug.cgi?id=1042933#c12

Thomas Blume <thomas.blume@xxxxxxxx> changed:

What |Removed |Added
----------------------------------------------------------------------------
Flags| |needinfo?(bpetkov@xxxxxxxx)

--- Comment #12 from Thomas Blume <thomas.blume@xxxxxxxx> ---
(In reply to Borislav Petkov from comment #11)
(In reply to Thomas Blume from comment #10)
You hit the nail.
The watchdog test takes ages.

I C.

The only question is now why it takes so long on SUSE whereas on
RH/Debian 10s are apparently sufficient.

Add more debugging output :-)

Here is a summary of my findings:

The code sets the watchdog timeout and then does a loop for pinging the
watchdog 5 times. On each iteration it sleeps for half of the watchdog timeout.
The watchdog_ping function opens the watchdog but doesn't call watchdog_close
after the ping is finished. Only the watchdog_close function contains the code
to stop the watchdog timer with the methods mentioned in commment#6.
This function is called only after the loop has finished which summarizes to a
minimum time of 5*watchdog_timeout/2 for the test.
Obviously this much too high for the timeout.

Boris, I'm wondering when exactly the watchdog timer starts counting down.
Is it when opening /dev/watchdog or when doing the ping (ioctl(watchdog_fd,
WDIOC_KEEPALIVE, 0) or something else?

--
You are receiving this mail because:
You are on the CC list for the bug.
< Previous Next >
References