On Aug 15, 2008, at 3:55 PM, Bert Wesarg wrote:
>> But I guess the cost is extremely low, and this isn't performance
>> critical
>> code, so having it check all the processors anyway is a safe thing
>> to do (in
>> case some machine someday *does* contain heterogeneous chips, all
>> running
>> Linux, which, with the advent of accelerators, I guess isn't *too*
>> far
>> fetched...). PLPA *almost* does this today, but could probably be
>> a bit
>> better about error checking.
> In the end, we need a valid mapping, i.e. processor_id <-> (socket_id,
> core_id). If we can't guarantee this, we should tell this.
I guess my point was that we might get mappings for *some* processors
and not others (certainly not on any of today's machines, but perhaps
that model might not be uncommon in the future...?). In this case,
perhaps plpa_have_topology_information() might return more than a bool:
- no topology information is available
- some topology information is available (i.e., using the plpa_()
functions will return info from what was available, but it doesn't
represent *all* processors)
- all topology information is availale
--
Jeff Squyres
Cisco Systems
|
![[Flag]](../../../../../community/mirrors/flag-us.gif){kind=small}