Messages in this thread | | | Date | Tue, 12 May 2009 17:43:26 +0200 | From | Andi Kleen <> | Subject | Re: [PATCH 3/3] x86: fix node_possible_map logic -v2 |
| |
> We actually have configurations that match both scenarios above. The > system is a blade-based system with 2 processor sockets per blade. > Memory is socket attached and each socket is in a unique PXM.
Jack, I think the interesting part would be: If you assign the CPUs to the next nearby node with memory. Do you see any unacceptable performance problems from that?
I'm sure the problem could be solved, but it would be quite some work, and that would be only worth spending if it's actually a significant benefit.
-Andi -- ak@linux.intel.com -- Speaking for myself only.
| |