|
|
|
|
|
|
|
|
|
|
xen-devel
Re: [Xen-devel] [RFC][PATCH 2/2] Add hcall to probe Xen heap
* Subrahmanian, Raj <raj.subrahmanian@xxxxxxxxxx> [2007-06-17 11:46]:
> Ryan,
> I applied these patches against 15200 and was able to boot on a 4-cell
> ES7000 with numa=on.
> Boots fine. I brought up a few VMs with some loads on them. That seems
> to work fine as well.
Great! Thanks for testing the patches out.
> Were there any other data points that you would like me to check out?
>
> I was able to see the node info from xm info
> node_to_cpu : node0:24-31
> node1:16-23
> node2:8-15
> node3:0-7
Yeah, in xm info, I'd like to see:
nr_cpus
nr_nodes
sockets_per_node
cores_per_socket
threads_per_core
I believe I've fixed up the sockets_per_node calculation, but the more
machines we test on the sooner we know for sure.
>
> If you could give me an updated set of patches against the tip, we can
> run our weekly tests with numa=on.
As Keir mentioned, you can still test with numa=on without the patches.
The new patches cover how the topology is exposed to userspace. In any
case, the latest versions are posted [1]here and [2]here.
1. http://lists.xensource.com/archives/html/xen-devel/2007-06/msg00298.html
2. http://lists.xensource.com/archives/html/xen-devel/2007-06/msg00299.html
--
Ryan Harper
Software Engineer; Linux Technology Center
IBM Corp., Austin, Tx
(512) 838-9253 T/L: 678-9253
ryanh@xxxxxxxxxx
_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel
|
|
|
|
|