On Fri, 2010-03-05 at 10:19 +0900, KAMEZAWA Hiroyuki wrote: > On Thu, 04 Mar 2010 12:06:54 -0500 > Lee Schermerhorn <lee.schermerhorn@xxxxxx> wrote: > > > >nid-04: > > > > > >* Isn't #define numa_mem numa_node a bit dangerous? Someone might use > > > numa_mem as a local variable name. Why not define it as a inline > > > function or at least a macro which takes argument. > > > > numa_mem and numa_node are the names of the per cpu variables, referenced > > by __this_cpu_read(). So, I suppose we can rename them both something like: > > percpu_numa_*. Would satisfy your concern? > > > > What do others think? > > > > Currently I've left them as numa_mem and numa_node. > > > > Could you add some documentation to Documentation/vm/numa ? > about > numa_node_id() > numa_mem_id() > topics on memory-less node > (cpu-less node) Hmmm. Good idea. I'll see what I can come up with. Thanks, Lee > > > Recently I see this kind of topics on list but I'm not sure whether > I catch the issues/changes correctly.... > > Thanks, > -Kame > -- To unsubscribe from this list: send the line "unsubscribe linux-numa" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
