WARNING - OLD ARCHIVES

This is an archived copy of the Xen.org mailing list, which we have preserved to ensure that existing links to archives are not broken. The live archive, which contains the latest emails, can be found at http://lists.xen.org/
   
 
 
Xen 
 
Home Products Support Community News
 
   
 

xen-bugs

[Xen-bugs] [Bug 1098] [version 3.1.0_15042-51] clocksource/0: Time went

To: xen-bugs@xxxxxxxxxxxxxxxxxxx
Subject: [Xen-bugs] [Bug 1098] [version 3.1.0_15042-51] clocksource/0: Time went backwards
From: bugzilla-daemon@xxxxxxxxxxxxxxxxxxx
Date: Wed, 31 Oct 2007 13:52:27 -0700
Delivery-date: Wed, 31 Oct 2007 13:52:53 -0700
Envelope-to: www-data@xxxxxxxxxxxxxxxxxx
In-reply-to: <bug-1098-3@xxxxxxxxxxxxxxxxxxxxxxxxxxx/bugzilla/>
List-help: <mailto:xen-bugs-request@lists.xensource.com?subject=help>
List-id: Xen Bugzilla <xen-bugs.lists.xensource.com>
List-post: <mailto:xen-bugs@lists.xensource.com>
List-subscribe: <http://lists.xensource.com/cgi-bin/mailman/listinfo/xen-bugs>, <mailto:xen-bugs-request@lists.xensource.com?subject=subscribe>
List-unsubscribe: <http://lists.xensource.com/cgi-bin/mailman/listinfo/xen-bugs>, <mailto:xen-bugs-request@lists.xensource.com?subject=unsubscribe>
Reply-to: bugs@xxxxxxxxxxxxxxxxxx
Sender: xen-bugs-bounces@xxxxxxxxxxxxxxxxxxx
http://bugzilla.xensource.com/bugzilla/show_bug.cgi?id=1098





------- Comment #4 from ian.marlier@xxxxxxxxxxxxxxxxxxx  2007-10-31 13:52 
-------
> This is my guess.
> 
> You should test with vcpu pin. It would also solve this problem.
> If my guess is correct, You shoud fix xen_clocksource_read() increment
> monotonically.
> currently vcpu migration causes time went backwards.
> 

I tested the system, following this advice, and there was no change in
behavior.

I pinned the CPU's this way:
offxen2:~ # xm vcpu-list
Name                              ID  VCPU   CPU State   Time(s) CPU Affinity
Domain-0                           0     0     0   -b-    1146.3 0
Domain-0                           0     1     1   r--      93.8 1
Domain-0                           0     2     2   -b-     234.5 2
Domain-0                           0     3     3   -b-    1022.0 3
SUDCX                              2     0     0   -b-      26.8 0
SUDCX                              2     1     1   -b-      27.3 1
SUDCX                              2     2     2   -b-      21.3 2
SUDCX                              2     3     3   -b-      15.9 3
alpharobot                         1     0     0   -b-      58.3 0
alpharobot                         1     1     1   -b-      55.2 1
mx1                                3     0     2   -b-      10.9 2
mx1                                3     1     3   -b-       7.5 3

And left the clocksource set to "xen" -- I rebooted the machine so it was in a
fully default state.
offxen2:~ # cat
/sys/devices/system/clocksource/clocksource0/current_clocksource 
xen 

The VMs started up without trouble.

As soon as I began burn-in testing in one of the domains, occasional messages
started appearing in the logs about time going backward.

>From the time that I got the burn-in tests going on all four of the domains,
the messages were appearing 100 or more times per minute.  After approximately
9 minutes of burnin testing in all four domains, the machine became
unresponsive at the console and to network, and had to be hard reset.

So, it appears that vcpu-pin does not have any effect on this issue.


-- 
Configure bugmail: 
http://bugzilla.xensource.com/bugzilla/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug, or are watching the assignee.

_______________________________________________
Xen-bugs mailing list
Xen-bugs@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-bugs