[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] [PATCH v2 3/4] sysctl: Add sysctl interface for querying PCI topology

>>> On 08.01.15 at 16:49, <boris.ostrovsky@xxxxxxxxxx> wrote:
> On 01/08/2015 04:50 AM, Jan Beulich wrote:
>>>>> On 07.01.15 at 18:55, <dario.faggioli@xxxxxxxxxx> wrote:
>>> There probably would not be too serious issues in converting everything
>>> to PXM, or adding duplicates, but I don't see the reason why we should
>>> do such a thing... Perhaps I'm missing what using PXM would actually buy
>>> us?
>> As long as those node IDs don't get stored persistently for future
>> use, using them (despite being a Xen internal representation)
>> ought to be fine. My concern really is that such Xen internals may
>> easily get mis-used in the tool stack, e.g. assuming that they won't
>> change across reboots (perhaps after a hypervisor update).
> Are PXMs guaranteed to be persistent across firmware upgrades? I don't 
> see anything about this in the APCI spec. The only semi-relevant 
> statement that I found was that PXM values are meaningless (i.e. they 
> are much like cookies).

Right, and I alluded to that earlier on. Still firmware updates are
presumable less frequent than hypervisor ones.

> (FWIW, Linux exposes nodeIDs in the same manner as Xen. In fact, Xen's 
> srat.c appears to be derived from Linux code.)

Not sure what the latter has to do with the former.


Xen-devel mailing list



Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.