[concurrency-interest] Multi-core testing, help with findings

David Holmes dcholmes at optusnet.com.au
Tue Dec 12 19:38:24 EST 2006


Correction:

> Both. The scheduling issue gives the busy-wait thread far too
> much CPU time.
> The fact that it is doing violatile accesses means that it is hitting the
> memory system in a way that interferes with the other threads too.

The spin loop does volatile reads which don't require fences/barriers, so it
shouldn't be causing interference.

It would be interesting to see how this performs without the spin-loop. It
might be all the fences induced by the worker threads as Bjorn suggested.

David Holmes



More information about the Concurrency-interest mailing list