Re: [Qemu-devel] [PATCH for-2.7 v6 2/2] spapr: implement query-hotpluggable-cpus callback

[Igor Mammedov]

On Mon, 11 Apr 2016 14:54:24 +1000
David Gibson <address@hidden> wrote:

> On Fri,  8 Apr 2016 13:29:56 +0200
> Igor Mammedov <address@hidden> 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 <address@hidden>
> > ---
> > it's only compile tested
> > v2:
> >  - s/qmp_query_hotpluggable_cpus/MachineClass->query_hotpluggable_cpus/ 
> > callback
> > ---
> >  hw/ppc/spapr.c | 33 +++++++++++++++++++++++++++++++++
> >  1 file changed, 33 insertions(+)
> > 
> > diff --git a/hw/ppc/spapr.c b/hw/ppc/spapr.c
> > index 760a42f..c38995e 100644
> > --- a/hw/ppc/spapr.c
> > +++ b/hw/ppc/spapr.c
> > @@ -66,6 +66,7 @@
> >  #include "hw/compat.h"
> >  #include "qemu/cutils.h"
> >  #include "hw/ppc/spapr_cpu_core.h"
> > +#include "qmp-commands.h"
> >  
> >  #include <libfdt.h>
> >  
> > @@ -2382,6 +2383,37 @@ static unsigned 
> > spapr_cpu_index_to_socket_id(unsigned cpu_index)
> >      return cpu_index / smp_threads / smp_cores;
> >  }
> >  
> > +static HotpluggableCPUList *spapr_query_hotpluggable_cpus(MachineState 
> > *machine)
> > +{
> > +    int i;
> > +    HotpluggableCPUList *head = NULL;
> > +    sPAPRMachineState *spapr = SPAPR_MACHINE(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);  
> 
> So, Bharata's latest core stuff is (as suggested by you and others)
> moving to using different derived types, rather than a cpu_model
> property, so this will need to change a bit.
Yep, I guess so.
It would be even better if Bharata took merged this patches
with his series and took care to make necessary changes to
this call back.

> 
> > +        cpu_item->vcpus_count = smp_threads;
> > +        cpu_props->has_core = true;
> > +        cpu_props->core = i;
> > +        /* TODO: add 'has_node/node' here to describe
> > +           to which node core belongs */
> > +
> > +        cpu_item->props = cpu_props;
> > +        if (spapr->cores[i]) {
> > +            cpu_item->has_qom_path = true;
> > +            cpu_item->qom_path = 
> > object_get_canonical_path(spapr->cores[i]);
> > +        }
> > +        list_item->value = cpu_item;
> > +        list_item->next = head;
> > +        head = list_item;
> > +    }
> > +    return head;
> > +}
> > +
> >  static void spapr_machine_class_init(ObjectClass *oc, void *data)
> >  {
> >      MachineClass *mc = MACHINE_CLASS(oc);
> > @@ -2412,6 +2444,7 @@ static void spapr_machine_class_init(ObjectClass *oc, 
> > void *data)
> >      hc->plug = spapr_machine_device_plug;
> >      hc->unplug = spapr_machine_device_unplug;
> >      mc->cpu_index_to_socket_id = spapr_cpu_index_to_socket_id;
> > +    mc->query_hotpluggable_cpus = spapr_query_hotpluggable_cpus;
> >  
> >      smc->dr_lmb_enabled = true;
> >      smc->dr_cpu_enabled = true;
> > -- 
> > 1.8.3.1
> >   
> 
>