Re: [Qemu-devel] [RFC PATCH v2 0/5] sPAPR: Fix migration when CPUs are removed in random order

[David Gibson <david@gibson.dropbear.id.au>]

[-- Attachment #1: Type: text/plain, Size: 1975 bytes --]

On Thu, Jul 07, 2016 at 06:04:10PM +0200, Greg Kurz wrote:
> On Thu,  7 Jul 2016 20:20:20 +0530
> Bharata B Rao <bharata@linux.vnet.ibm.com> wrote:
> 
> > device_add/del based CPU hotplug and unplug support is upstream for
> > sPAPR PowerPC and is under development for x86. Both of these will
> > support CPU device removal in random order (and not necessarily in LIFO
> > order). Random order removal will result in holes in cpu_index range
> > which causes migration to fail. This needs fixes in both generic code
> > as well as arch specific code.
> > 
> > - CPUState::stable_cpu_id is newly introduced and used as instance_id when
> >   registering CPU devices using vmstate_register. stable_cpu_id is set by the
> >   target machine code. To support forward migration, as per Igor's
> >   suggestion, this needs to be done conditionally based on machine type
> >   version.
> > - From pseries-2.7 onwards, we start using stable_cpu_id for migration as
> >   well as in XICS code.
> > 
> > vmstate registration calls are moved to cpu_common_realizefn and newly
> > introduced cpu_common_unrealizefn.
> > 
> > This patchset depends on Greg Kurz's patchset where among other things,
> > he is deriving cpu_dt_it (which is stable_cpu_id for pseries-2.7 onwards)
> > based on core-id and hence is based on ppc-vcpu-dt-id-rework branch of his
> > tree.
> > 
> 
> I'm not very comfortable with this. Shouldn't it be the other way round
> actually: cpu_dt_id depending on stable_cpu_id ?
> 
> I think we're missing something like a stable_core_id.

The core-id is already stable.

Deriving the stable vcpu id from the core id is correct.

cpu_dt_id should actually go away, and we should just use
stable_cpu_id when creating the device tree.

-- 
David Gibson			| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au	| minimalist, thank you.  NOT _the_ _other_
				| _way_ _around_!
http://www.ozlabs.org/~dgibson

[-- Attachment #2: signature.asc --]
[-- Type: application/pgp-signature, Size: 819 bytes --]