Open MPI logo

Hardware Locality Development Mailing List Archives

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

Subject: Re: [hwloc-devel] hwloc_alloc_membind
From: Karl Napf (karl.b.napf_at_[hidden])
Date: 2012-02-24 08:44:55


Thanks a lot, Samuel. BTW the documentation says that
hwloc_alloc_membind and hwloc_alloc_membind_nodeset return -1 on
error, but this should be NULL.

Regards
Karl

2012/2/24 Samuel Thibault <samuel.thibault_at_[hidden]>:
> Karl Napf, le Fri 24 Feb 2012 13:04:58 +0100, a écrit :
>> What surprises me is that the result of the call to
>> hwloc_fix_membind_cpuset in line 534 of bind.c is negated, even though
>> hwloc_fix_membind_cpuset returns 0 on success. Might this be a bug?
>
> Oops, indeed.
>
>> 2. In another use case I want to allocate memory using the first-touch
>> policy (HWLOC_MEMBIND_FIRSTTOUCH). Does the cpuset passed to
>> hwloc_alloc_membind have any meaning in this case, since the memory is
>> not immediately bound to a node?
>
> It has a meaning when there is not enough memory on the node. Allocation
> may then be done in some node of the given cpuset.  This kind of precise
> detail on the behavior however vary over OSes.  Linux does not take it
> into account for instance.
>
> Samuel
> _______________________________________________
> hwloc-devel mailing list
> hwloc-devel_at_[hidden]
> http://www.open-mpi.org/mailman/listinfo.cgi/hwloc-devel