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 went backwards

On Sat, 11 Mar 2006, Keir Fraser wrote:
> On 10 Mar 2006, at 21:57, Rik van Riel wrote:
> 
> > stolen + blocked is larger than delta_cpu all the time:
> > 
> > Timer ISR/0: stolen 1636284 + blocked 109702879 > delta_cpu 108006671
> > printk: 33 messages suppressed.
> 
> Is sched_time > (processed_system_time + delta)? That would be the real
> killer.

Very frequently.  By significant amounts, too...

No, I have no idea why...

Timer ISR/0: stolen 1921037 + blocked 79704210 > delta_cpu 79978775
Timer ISR/0: sched_time 429620376038547 > delta 4006385 + cpu 
processed_system_time 429620356036193
Timer ISR/0: sched_time 429620710119143 > delta_cpu 206088604 + global 
processed_system_time 429620500036193
Timer ISR/0: sched_time 429621780039343 > delta 404013043 + cpu 
processed_system_time 429621360036193
Timer ISR/0: sched_time 429622284012371 > delta_cpu -11948 + global 
processed_system_time 429622280036193

> Because we only account stolen/blocked/running time in jiffy-sized chunks,
> it's possible for the three deltas to be out of sync with each other by a
> jiffy.

But we round everything down, so while it is possible for the
sum of rounded down stolen/blocked/running/idle to be way less
than the total, I don't see how the sum of these rounded down
values could ever be more than the total (which is only rounded
off once)...

-- 
All Rights Reversed

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