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-devel

RE: [Xen-devel] Time stopped

To: "David F Barrera" <dfbp@xxxxxxxxxx>
Subject: RE: [Xen-devel] Time stopped
From: "Ian Pratt" <m+Ian.Pratt@xxxxxxxxxxxx>
Date: Thu, 13 Oct 2005 20:21:01 +0100
Cc: xen-devel <xen-devel@xxxxxxxxxxxxxxxxxxx>
Delivery-date: Thu, 13 Oct 2005 19:18:18 +0000
Envelope-to: www-data@xxxxxxxxxxxxxxxxxxx
List-help: <mailto:xen-devel-request@lists.xensource.com?subject=help>
List-id: Xen developer discussion <xen-devel.lists.xensource.com>
List-post: <mailto:xen-devel@lists.xensource.com>
List-subscribe: <http://lists.xensource.com/cgi-bin/mailman/listinfo/xen-devel>, <mailto:xen-devel-request@lists.xensource.com?subject=subscribe>
List-unsubscribe: <http://lists.xensource.com/cgi-bin/mailman/listinfo/xen-devel>, <mailto:xen-devel-request@lists.xensource.com?subject=unsubscribe>
Sender: xen-devel-bounces@xxxxxxxxxxxxxxxxxxx
Thread-index: AcXQKUQ+Bzp6Y7bqR3S3ZhXbBeAP4AAAMH2A
Thread-topic: [Xen-devel] Time stopped
 

> Recreated the situation on non-PAE kernel on the same machine.

As I suspected.
 
> On Thu, 2005-10-13 at 18:29 +0100, Ian Pratt wrote:
> 
> > 
> > When the machine is in this state, what interrupt rate does 
> > /proc/interrupts show?
> x335b:~ # cat /proc/interrupts

You'll need to do this multiple times e.g. 10 seconds appart for us to
figure out the rate.

> > If you do a "sleep 1" how long does it sleep for?
> In this case, it slept for about 3 real minutes.

Interesting.
 
> > I presume both dom0 and the domU experience the time slow down?
> Correct. 
> > 
> > If you run something in the background that burns CPU does 
> it make a 
> > difference?
> I started LTP on dom0 ( and can see console outpu) and it 
> makes no difference.

I think a "while true; do true; done" might be more convincing, but LTP
is probably good enough.

> > In this case, what does xm list show as regards the CPU 
> time consumed 
> > by the domains?
> x335b:/tmp/ltp-full-20050804 # date
> Thu Oct 13 13:47:59 CDT 2005
> x335b:/tmp/ltp-full-20050804 # xm list
> Name              ID  Mem(MiB)  CPU  VCPUs  State   Time(s)
> Domain-0           0       495    -      4  r-----     96.5
> 11_create_0       73        16    3      1  -b----      0.3
> x335b:/tmp/ltp-full-20050804 # xm list
> Name              ID  Mem(MiB)  CPU  VCPUs  State   Time(s)
> Domain-0           0       495    -      4  r-----     96.5
> 11_create_0       73        16    3      1  -b----      0.3
> x335b:/tmp/ltp-full-20050804 # xm list
> Name              ID  Mem(MiB)  CPU  VCPUs  State   Time(s)
> Domain-0           0       495    -      4  r-----     96.5
> 11_create_0       73        16    3      1  -b----      0.3
> x335b:/tmp/ltp-full-20050804 #
> x335b:/tmp/ltp-full-20050804 # date
> Thu Oct 13 13:47:59 CDT 2005

I didn't expect this. Need to think some.

> > Can you narrow down which of the xm-tests actually 
> provoke's the bug?

Can you repro using just this test?

If you had a serial console and perfc=y enabled it would be interesting
to see the real interrupt rate from the xen debug console.

Ian

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

<Prev in Thread] Current Thread [Next in Thread>