Re: Xen spinlock questions

["Jan Beulich" <jbeulich@novell.com>]

>>> Keir Fraser <keir.fraser@eu.citrix.com> 12.08.08 19:00 >>>
>On 12/8/08 17:33, "Jeremy Fitzhardinge" <jeremy@goop.org> wrote:
>
>>>> That seems reasonable.  In this use-case, it's quite likely that if the
>>>> poll is interrupted by event delivery, on return it will find that the
>>>> spinlock is now free and never re-enter the poll.
>>>>     
>>> 
>>> Attached is a new version of the patch which clears the vcpu from poll_mask
>>> when it is unblocked for any reason. Jan: please can you give this one a
>>> spin if you get time.
>>>   
>> 
>> Forgot to attach?
>
>Sorry, attached now.

I can't really explain the results of testing with this version of the patch:
While the number of false wakeups got further reduced by somewhat
less than 20%, both time spent in the kernel and total execution time
went up (8% and 4% respectively) compared to my original (and from
all I can tell worse) version of the patch. Nothing else changed as far as
I'm aware.

Jan