Re: [libvirt] [PATCH] nodeinfo: Don't fail on non-contiguous NUMA topologies

2013-07-12 Thread Peter Krempa
On 07/11/13 18:08, Daniel P. Berrange wrote: On Thu, Jul 11, 2013 at 04:09:47PM +0200, Peter Krempa wrote: From: hejia hejia jiaker...@gmail.com nodeGetFreeMemory and nodeGetCellsFreeMemory assumed that the NUMA nodes are contiguous and starting from 0. Unfortunately there are machines that

[libvirt] [PATCH] nodeinfo: Don't fail on non-contiguous NUMA topologies

2013-07-11 Thread Peter Krempa
From: hejia hejia jiaker...@gmail.com nodeGetFreeMemory and nodeGetCellsFreeMemory assumed that the NUMA nodes are contiguous and starting from 0. Unfortunately there are machines that don't match this assumption: available: 1 nodes (1) node 1 cpus: 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 node 1

Re: [libvirt] [PATCH] nodeinfo: Don't fail on non-contiguous NUMA topologies

2013-07-11 Thread Daniel P. Berrange
On Thu, Jul 11, 2013 at 04:09:47PM +0200, Peter Krempa wrote: From: hejia hejia jiaker...@gmail.com nodeGetFreeMemory and nodeGetCellsFreeMemory assumed that the NUMA nodes are contiguous and starting from 0. Unfortunately there are machines that don't match this assumption: available: 1