On Wed, 16 Mar 2016 10:49:41 +0530 Bharata B Rao <bhar...@linux.vnet.ibm.com> wrote:
> On Tue, Mar 15, 2016 at 02:24:08PM +0100, Igor Mammedov wrote: > > it returns a list of present/possible to hotplug CPU > > objects with a list of properties to use with > > device_add. > > > > in spapr case returned list would looks like: > > -> { "execute": "query-hotpluggable-cpus" } > > <- {"return": [ > > { "props": { "core": 1 }, "type": "spapr-cpu-core", > > "vcpus-count": 2 }, > > { "props": { "core": 0 }, "type": "spapr-cpu-core", > > "vcpus-count": 2, > > "qom-path": "/machine/unattached/device[0]"} > > ]}' > > > > TODO: > > add 'node' property for core <-> numa node mapping > > > > Signed-off-by: Igor Mammedov <imamm...@redhat.com> > > --- > > it's only compile tested > > --- > > hw/ppc/spapr.c | 32 ++++++++++++++++++++++++++++++++ > > 1 file changed, 32 insertions(+) > > > > diff --git a/hw/ppc/spapr.c b/hw/ppc/spapr.c > > index b1e9ba2..e1ce983 100644 > > --- a/hw/ppc/spapr.c > > +++ b/hw/ppc/spapr.c > > @@ -65,6 +65,7 @@ > > #include "hw/compat.h" > > #include "qemu-common.h" > > #include "hw/ppc/spapr_cpu_core.h" > > +#include "qmp-commands.h" > > > > #include <libfdt.h> > > > > @@ -2399,6 +2400,37 @@ static unsigned > > spapr_cpu_index_to_socket_id(unsigned cpu_index) > > return cpu_index / smp_threads / smp_cores; > > } > > > > +HotpluggableCPUList *qmp_query_hotpluggable_cpus(Error **errp) > > +{ > > + int i; > > + HotpluggableCPUList *head = NULL; > > + sPAPRMachineState *spapr = SPAPR_MACHINE(qdev_get_machine()); > > + int spapr_max_cores = max_cpus / smp_threads; > > + > > + for (i = 0; i < spapr_max_cores; i++) { > > + HotpluggableCPUList *list_item = g_new0(typeof(*list_item), 1); > > + HotpluggableCPU *cpu_item = g_new0(typeof(*cpu_item), 1); > > + CpuInstanceProperties *cpu_props = g_new0(typeof(*cpu_props), 1); > > + > > + cpu_item->type = g_strdup(TYPE_SPAPR_CPU_CORE); > > + cpu_item->vcpus_count = smp_threads; > > Shouldn't this be fetched from "threads" property of the core device > instead of directly using smp_threads ? But again, what that would mean > for not-yet-plugged in cores and how to get that for them is a question. my understanding is that libvirt wants to know that info for unplugged CPUs as well and machine model + core(cpu_model) combo would fix it at machine startup time for current use case. > > Regards, > Bharata. >