Messages in this thread | | | Date | Thu, 1 Aug 2013 12:12:56 +0200 | From | Peter Zijlstra <> | Subject | Re: [PATCH RFC 1/2] qspinlock: Introducing a 4-byte queue spinlock implementation |
| |
On Thu, Aug 01, 2013 at 03:41:33PM +0530, Raghavendra K T wrote: > It is exactly 16k-1 not 15k > That is because CPU_CODE of 1 to 16k represents cpu 0..16k-1
From what I know big systems are usually build with power-of-two factors. Although I suppose with a ring fabric you could have an arbitrary number of nodes.
Anyway, I've heard SGI talk about 4K cpu systems, 8K cpu systems and 16K cpu systems, I've not heard them talk about 16K-n systems.
Also, as in other parts of the reply I send, this limitation seems completely unnecessary.
| |