Charm++ parallel programming system

[Jim Phillips <jim AT ks.uiuc.edu>]

Can we automate including it when it is needed somehow, or at least take 
it out so the typical case works out of the box?

-Jim

On Tue, 8 Feb 2011, Abhinav S Bhatele wrote:

> rca is used to get physical coordinates for each physical node using the
> rca_get_meshcoord call.
> This need not be linked in since the code is not turned on by default.
>
> - Abhinav
>
>
> On Tue, Feb 8, 2011 at 2:35 PM, Jim Phillips 
> <jim AT ks.uiuc.edu>
>  wrote:
>
> > Thanks.  Anybody know what -lrca does?  -Jim
> >
> >
> > On Tue, 8 Feb 2011, Abhinav S Bhatele wrote:
> >
> >  Hi Viraj,
> > > I downloaded the development version of charm and NAMD and they compiled
> > > just fine.
> > > Here are the steps I followed:
> > >
> > > module swap PrgEnv-pgi PrgEnv-gnu
> > >
> > > Compile charm as:
> > > ./build charm++ mpi-crayxt -j8 -g
> > >
> > > Compile namd as:
> > > ./config CRAY-XT-g++ --without-tcl --without-fftw
> > >
> > > To get NAMD to link, I had to remove -lrca from
> > > charm/src/arch/mpi-crayxt/conv-mach.sh
> > >
> > > And then I successfully ran NAMD on 24 cores.
> > >
> > > - Abhinav
> > >
> > >
> > > On Wed, Jan 26, 2011 at 2:52 PM, Viraj Paropkari 
> > > <vrparopkari AT lbl.gov
> > > > wrote:
> > >
> > >  NAMD 2.7 and charm 6.2.2
> > > > On Tue, Jan 25, 2011 at 5:16 PM, Phil Miller 
> > > > <mille121 AT illinois.edu
> > > > > wrote:
> > > >
> > > >  On Tue, Jan 25, 2011 at 19:09, Viraj Paropkari 
> > > > <vrparopkari AT lbl.gov>
> > > > > wrote:
> > > > >
> > > > > > Hi Abhinav,
> > > > > >                     Here is what I tried..
> > > > > >
> > > > > > I had trouble to build the latest version of NAMD 2.7 on Hopper II, so
> > > > > > I
> > > > > > built a bit older cvs version of NAMD on it instead. This cvs version
> > > > > was
> > > > >
> > > > > > built in the exactly same way as building cray-xt5 on Hopper Phase I,
> > > > > and it
> > > > >
> > > > > > runs fine. The problem is NAMD 2.7 code compiles fine, but the binary
> > > > > > doesn't run because the Charm++ couldn't get the multi-thread support
> > > > > level
> > > > >
> > > > > > correctly (do not have log of exact error messages right now). This was
> > > > > with
> > > > >
> > > > > > latest version of charm++...
> > > > > >
> > > > > > If you want more information of exact build line etc , let me know.
> > > > >
> > > > > This is exactly what Abhinav was asking for. Also, when you say "the
> > > > > latest version of charm++", do you mean the 6.2.2 release, or
> > > > > something gotten from Git? If the latter, what commit id?
> > > > >
> > > > >  Also let me know if there is an update for your effort on XE6.
> > > > > > Thanks.
> > > > > >
> > > > > >
> > > > > >
> > > > > >
> > > > > > On Tue, Jan 25, 2011 at 4:48 PM, Abhinav S Bhatele <
> > > > > bhatele AT illinois.edu>
> > > > >
> > > > > > wrote:
> > > > > >
> > > > > > > What charm and NAMD build did you try? And what error messages are you
> > > > > > > seeing when building Charm?
> > > > > > > - A
> > > > > > >
> > > > > > > On Tue, Jan 25, 2011 at 6:45 PM, Viraj Paropkari 
> > > > > > > <vrparopkari AT lbl.gov
> > > > > > > >
> > > > > > > wrote:
> > > > > > >
> > > > > > > > Hi Abhinav,
> > > > > > > >                     Any updates on this one ? I also wanted to ask if
> > > > > > > > charm++ has been ported on XE6 ? We are seeing some issues while
> > > > > > > building
> > > > >
> > > > > > NAMD 2.7 on XE6 which seem to be coming from charm++. Error messages
> > > > > > > >
> > > > > > > are
> > > > >
> > > > > > about wrong level of multi thread support. Although namd2 binary from
> > > > > > > >
> > > > > > > CVS
> > > > >
> > > > > > version works.
> > > > > > > > Thanks.
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > > On Thu, Jan 6, 2011 at 5:06 PM, Abhinav S Bhatele <
> > > > > > > bhatele AT illinois.edu>
> > > > >
> > > > > > wrote:
> > > > > > > > > Hi Viraj,
> > > > > > > > > As far as I know Cray has tried NAMD on some XE6's before they sold
> > > > > > > > > them. We haven't tried running it yet but I am planning to in the
> > > > > > > > next few
> > > > >
> > > > > > weeks.
> > > > > > > > > - Abhinav
> > > > > > > > >
> > > > > > > > > On Thu, Jan 6, 2011 at 7:01 PM, Viraj Paropkari <
> > > > > > > > > vrparopkari AT lbl.gov
> > > > > >  wrote:
> > > > > > > > > > Hi PPLers,
> > > > > > > > > >                   Just wondering if group has tried running NAMD
> > > > > > > > > 2.7
> > > > >
> > > > > > on Cray XE6 ?
> > > > > > > > > >                  Thanks.
> > > > > > > > > >
> > > > > > > > > > --
> > > > > > > > > > ------------------------------------------------------------
> > > > > > > > > > Viraj R. Paropkari
> > > > > > > > > >
> > > > > > > > > > HPC Engineer
> > > > > > > > > > NERSC, Lawrence Berkeley National Lab
> > > > > > > > > > One Cyclotron Road, MS 943-256
> > > > > > > > > > Berkeley, CA  94720-8043
> > > > > > > > > > 510.486.4520 voice
> > > > > > > > > > 510.486.4316 fax
> > > > > > > > > > VRParopkari AT lbl.gov
> > > > > > > > > > ------------------------------------------------------------
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > >
> > > > > > > > > > _______________________________________________
> > > > > > > > > > ppl mailing list
> > > > > > > > > > ppl AT cs.uiuc.edu
> > > > > > > > > > http://lists.cs.uiuc.edu/mailman/listinfo/ppl
> > > > > > > > >
> > > > > > > > >
> > > > > > > > > --
> > > > > > > > > Abhinav S Bhatele, Ph. D.
> > > > > > > > > Webpage: www.bhatele.org
> > > > > > > >
> > > > > > > >
> > > > > > > >
> > > > > > > > --
> > > > > > > > ------------------------------------------------------------
> > > > > > > > Viraj R. Paropkari
> > > > > > > >
> > > > > > > > HPC Engineer
> > > > > > > > NERSC, Lawrence Berkeley National Lab
> > > > > > > > One Cyclotron Road, MS 943-256
> > > > > > > > Berkeley, CA  94720-8043
> > > > > > > > 510.486.4520 voice
> > > > > > > > 510.486.4316 fax
> > > > > > > > VRParopkari AT lbl.gov
> > > > > > > > ------------------------------------------------------------
> > > > > > >
> > > > > > >
> > > > > > > --
> > > > > > > Abhinav S Bhatele, Ph. D.
> > > > > > > Webpage: www.bhatele.org
> > > > > >
> > > > > >
> > > > > >
> > > > > > --
> > > > > > ------------------------------------------------------------
> > > > > > Viraj R. Paropkari
> > > > > >
> > > > > > HPC Engineer
> > > > > > NERSC, Lawrence Berkeley National Lab
> > > > > > One Cyclotron Road, MS 943-256
> > > > > > Berkeley, CA  94720-8043
> > > > > > 510.486.4520 voice
> > > > > > 510.486.4316 fax
> > > > > > VRParopkari AT lbl.gov
> > > > > > ------------------------------------------------------------
> > > > > >
> > > > > >
> > > > > >
> > > > > > _______________________________________________
> > > > > > ppl mailing list
> > > > > > ppl AT cs.uiuc.edu
> > > > > > http://lists.cs.uiuc.edu/mailman/listinfo/ppl
> > > >
> > > >
> > > > --
> > > > ------------------------------------------------------------
> > > > Viraj R. Paropkari
> > > >
> > > > HPC Engineer
> > > > NERSC, Lawrence Berkeley National Lab
> > > > One Cyclotron Road, MS 943-256
> > > > Berkeley, CA  94720-8043
> > > > 510.486.4520 voice
> > > > 510.486.4316 fax
> > > > VRParopkari AT lbl.gov
> > > > ------------------------------------------------------------
> > >
> > > --
> > > Abhinav S Bhatele, Ph. D.
> > > Webpage: www.bhatele.org
>
>
> --
> Abhinav S Bhatele, Ph. D.
> Webpage: www.bhatele.org