On Wed, Sep 17, 2014 at 02:19:05PM +0400, Andrey Korolyov wrote: > On Wed, Sep 17, 2014 at 2:00 PM, Tang Chen <tangc...@cn.fujitsu.com> wrote: > > Add Andrey Korolyov <and...@xdel.ru> > > > > On 09/17/2014 04:32 PM, Hu Tao wrote: > >> > >> On Tue, Sep 16, 2014 at 06:39:15PM +0800, zhanghailiang wrote: > >>> > >>> If we do not configure numa option, memory hotplug should work as well. > >>> It should not depend on numa option. > >>> > >>> Steps to reproduce: > >>> (1) Start VM: qemu-kvm -m 1024,slots=4,maxmem=8G > >>> (2) Hotplug memory > >>> It will fail and reports: > >>> "'DIMM property node has value 0' which exceeds the number of numa nodes: > >>> 0" > >>> > >> I rememberd Tang Chen had a patch for this bug, this is what Andrey > >> suggested: > >> > >> I thnk that there should be no > >> cases when dimm is plugged (and check from patch is fired up) without > >> actually populated NUMA, because not every OS will workaround this by > >> faking the node. > > > > > > According to Andrey Korolyov <and...@xdel.ru>, memory hotplug should not > > work without SRAT. So maybe forcing to create a NUMA node and SRAT will be > > better idea. > > > > I'm have been working on it. > > > > Thanks. > > > > > >> > >> https://lists.nongnu.org/archive/html/qemu-devel/2014-08/msg04587.html > >> > >> Have you tested this patch with Windows guest? > >> > >> Regards, > >> Hu > >> > >> > > > > Thanks, is there will be a place to guard against misconfiguration in > the dimm properties for NUMA too? For example, right now I may specify > just one node for a topo and assign dimms to more than one node, which > will pass argument checks and will result in memory allocation errors > in guest.
Yes, there is, in function pc_dimm_realize(). Regards, Hu