Open MPI logo

Hardware Locality Development Mailing List Archives

  |   Home   |   Support   |   FAQ   |   all Hardware Locality Development mailing list

Subject: Re: [hwloc-devel] dynamic cpuset_t?
From: Fawzi Mohamed (fawzi_at_[hidden])
Date: 2009-09-29 14:57:24


On 29-set-09, at 20:51, Fawzi Mohamed wrote:

>
> On 29-set-09, at 20:43, Samuel Thibault wrote:
>
>> Fawzi Mohamed, le Tue 29 Sep 2009 20:39:02 +0200, a écrit :
>>> It comes down to what you want to have, if you think you might
>>> want to
>>> go the sparse full granularity way then indeed alloc/copy/free
>>> should
>>> be added
>>
>> Yes, that's my point: do we really want it?
>
> I am more for avoiding it, if needed by using granularity, and
> having extracting functions that might build the cpuset on the fly I
> think that most of advantages of allocating/freeing/copying can be
> achieved without the annoying free/alloc.

mmh thinking more I found a reason to add it: if the conversion to/
buildup of the "native" cpuset costly, and this platform has alloc/
free for its native cpuset...
then having alloc/free would allow one to use directly the native
cpuset on that platform (or embed it in the cpuset).
So that is probably a good reason to bit the bullet and add it.

Fawzi