Re: A clocksource question

[Joanna Rutkowska <joanna@invisiblethingslab.com>]

[-- Attachment #1.1: Type: text/plain, Size: 1623 bytes --]

On 03/12/2010 01:38 AM, Joanna Rutkowska wrote:
> On 03/12/2010 01:49 AM, Jeremy Fitzhardinge wrote:
>> On 03/11/2010 04:27 PM, Joanna Rutkowska wrote:
>>> I think that the important clue is this message appearing in my dmesg
>>> (in Dom0 and also in DomUs):
>>>
>>> hrtimer: interrupt too slow, forcing clock min delta to 540150561 ns
>>>
>>> This is almost 0.5s (!) and I think this might explain my kbd hiccup. I
>>> wrote that I feel it every 10s or so, but when I was playing on my
>>> system without jiffies setting today, I saw this hiccup occurring much
>>> more often; in fact it was more of a "slow keyboard/system" feeling than
>>> a hiccup".
>>>    
>>
>> Yes, its definitely a good clue.  It could point to a Xen scheduler
>> issue (I assume you don't have any other busy domains when this is going
>> on), but that wouldn't explain why it works OK with jiffies.  Perhaps
>> there's a problem with programming the timer so that sometimes it gets
>> missed and the system doesn't get a kick until some subsequent event
>> (whereas with clocksource=jiffies there'll always be a timer in the next
>> 100ms).
>>
> One detail: when I boot the system and there is only Dom0 running,
> everything seems ok. Only after I start at least one VM the problem
> becomes noticeable. But the other VM(s) is pretty much idle (as xentop
> shows).
> 
> When I kill all the other VMs, so that Dom0 is alone again, the problem
> does *not* seem to go away, strangely...
> 
Today I upgraded to xen 3.4.3-rc3 and tried the 2.6.32 kernel (the
Micheal Young's one) -- the same problem :(

joanna.


[-- Attachment #1.2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 226 bytes --]

[-- Attachment #2: Type: text/plain, Size: 138 bytes --]

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