On 01/17/13 17:56, Daniel P. Berrange wrote:
On Thu, Jan 17, 2013 at 05:14:30PM +0100, Peter Krempa wrote:
On 01/17/13 16:43, Daniel P. Berrange wrote: VDSM uses the <topology> element (subelement of cpu) in the capabilities and multiplies the numbers by the number of numa nodes. As the data there is taken from nodeinfo, this breaks on some systems.
If it is merely looking at the current CPU <topology> then I see no reason why we can't make enough information available in the NUMA <topology> to let it do the right thing on all systems.
We definitely can. I'm not against it, I just think it's a pretty heavy hammer for the problem. But as long as it provides reasonable data I don't care. The thing I care about is providing reasonable data so that the AMD Bulldozer CPU can be detected in either way depending on the choice of the management app.
Regards, Daniel