Re: [hwloc-devel] hwloc version number

Errr... I got the patch slightly wrong. The one added line should include the ".$HWLOC_RELEASE_VERSION" at the end. You get the idea... On Oct 1, 2014, at 8:50 PM, Jeff Squyres (jsquyres) wrote: > We just changed the version numbering scheme over in OMPI to always be >

[hwloc-devel] hwloc version number

We just changed the version numbering scheme over in OMPI to always be "a.b.c", even if c==0. Do we want the same thing in hwloc? Right now, hwloc uses the old OMPI versioning scheme -- it's "a.b.c", except when c==0, and then the version is "a.b". It's a pretty simple change: diff --git

Re: [OMPI devel] Neighbor collectives with periodic Cartesian topologies of size one

I should have been clearer. This changes how the topo components are expected to work. With this change they now MUST provide the c_topo to the communicator before coll_select. The change is good but may be unexpected for any vendor topo components out there (which I doubt there are any). -Nathan

Re: [OMPI devel] Neighbor collectives with periodic Cartesian topologies of size one

I see no change in the topo interface in any of the patches attached to this thread. Is there any other patch related to this discussion? George. > On Oct 1, 2014, at 14:52, Jeff Squyres (jsquyres) wrote: > >> On Oct 1, 2014, at 6:48 AM, Gilles Gouaillardet >>

Re: [OMPI devel] Neighbor collectives with periodic Cartesian topologies of size one

On Oct 1, 2014, at 6:48 AM, Gilles Gouaillardet wrote: > oops, i missed graph and dist graph :-( > > the attached patch fixes this > /* it was straightforward for graph and a bit more challenging for dist > graph */ > > i am fine for Nathan's patch for v1.8 > /*

Re: [OMPI devel] Neighbor collectives with periodic Cartesian topologies of size one

oops, i missed graph and dist graph :-( the attached patch fixes this /* it was straightforward for graph and a bit more challenging for dist graph */ i am fine for Nathan's patch for v1.8 /* v1.8 and trunk have already a different way to handle a memory leak */ /* the attached patch is vs v1.8

Re: [OMPI devel] MPI_Comm_spawn crashes with the openib btl

Thanks Ralph ! it did fix the problem Cheers, Gilles On 2014/10/01 3:04, Ralph Castain wrote: > I fixed this in r32818 - the components shouldn't be passing back success if > the requested info isn't found. Hope that fixes the problem. > > > On Sep 30, 2014, at 1:54 AM, Gilles Gouaillardet >