On Mon, Mar 06, 2006 at 08:07:26PM +0100, Jiri Benc wrote: > On Mon, 6 Mar 2006 10:49:46 -0800, Jouni Malinen wrote: > > The reason for this optimization was in even high-end CPUs starting to > > run out of resources when running one radio with 2007 "virtual STAs",
> Yes, I'm aware of that. But I'm afraid that hard-wired need for STA > interfaces to have incremental MAC addresses is not acceptable - and > this fact alone means degrading of performance in case of 2000+ STAs. Since this is not needed for most normal use cases, this is probably fine. However, being able to test with large number of STAs is very useful test case for AP functionality (number of APs crash if you try to associate that many STAs..) and performance. One approach could be to add a hash table for address to netdev mapping. If that is not easily doable, a (hopefully) small patch could of course be maintained separately, but I would prefer to see this operation mode as something that is supported in the stack by default. -- Jouni Malinen PGP id EFC895FA - To unsubscribe from this list: send the line "unsubscribe netdev" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
