2012/7/27 K. Frank <[email protected]>:
> Hi Kai!
>
> On Fri, Jul 27, 2012 at 4:09 PM, Kai Tietz <[email protected]> wrote:
>> So, I made spinlocks "fair" by revision 5274.  This means that no
>> threads get *lost* on scheduling, if lock is requested.
>
> This is purely a matter of multi-threading curiosity on my part.
>
> How much additional cost (if any) do you think fairness adds?
>
>> Thanks in advance,
>> Kai
>
> As always, thanks for all of your great work.
>
>
> K. Frank

Well, the additional cost in terms of computation we can ignore.
Right now I have implemented it that way, that it uses a call to
Sleep(0) on waiting for locking.  This can introduce a speed-penalty
as task-switching on windows via sleep takes some time.

Regards,
Kai

------------------------------------------------------------------------------
Live Security Virtual Conference
Exclusive live event will cover all the ways today's security and 
threat landscape has changed and how IT managers can respond. Discussions 
will include endpoint security, mobile security and the latest in malware 
threats. http://www.accelacomm.com/jaw/sfrnl04242012/114/50122263/
_______________________________________________
Mingw-w64-public mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/mingw-w64-public

Reply via email to