[Xen-devel] Re: A question about changeset 20621:f9392f6eda79 and Discontinuous online node

[Keir Fraser <keir.fraser@xxxxxxxxxxxxx>]

On 06/01/2010 08:14, "Jiang, Yunhong" <yunhong.jiang@xxxxxxxxx> wrote:

> This continuous assumption apply to CPU side also. Currently nr_cpus is
> returned as num_online_cpus(), this will cause issue if some of cpu is
> offlined.
> 
> I'm considering if we can pass this dis-continuous information to user space
> too, but that requires change this sysctl interface. The worse is, even if we
> can change this interface, we may run out of the 128 byte limitation for
> xen_sysctl hypercall if we change the NR_CPUS == 128 in future (currently the
> struct xen_sysctl_physinfo is 104 byte already).
> 
> I'd get some input from you guys and community before I try to fix this issue,
> any suggestion?

I think sufficient info is already made available by sysctl: cpu_to_node[]
info is provided by physinfo command; cpu_to_core[] and cpu_to_socket[] by
get_cputopo command. In both cases holes due to offline cpus are filled with
a special indicator value of ~0.

In the case of the node_to_cpu list returned by xc.physinfo(), it already
accounts for holes in node space by having empty lists of cpus for those
non-existent or empty nodes. Should just be a case of having the xend python
code understand that as necessary?

 -- Keir



_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel