This commit produces warnings in the output of oshmem_info:
OSHMEM Fort shmem.fh: yes
**
*** DEVELOPER WARNING: A field in ompi_info output is too long and
*** will appear poorly in the prettyprint output.
***
*** Value:
More comments on this commit:
- The Fortran, Java, and C++ MPI examples are now no longer build by default.
Er... what happened there, and why?
- Why are the oshmem examples in a separate target? The point of the previous
makefile was that "make" (or "make all") would build all relevant examp
Final comment on this commit: please set svn:ignore on your commits properly:
❯❯❯ svn st .
? hello_shmemfh
? ring_shmem
? ring_shmemfh
On Nov 11, 2013, at 2:49 PM, "Jeff Squyres (jsquyres)"
wrote:
> More comments on this commit:
>
> - The Fortran, Java, and C++ MPI examples
On 11/11/13 12:49 PM, "Jeff Squyres (jsquyres)" wrote:
>More comments on this commit:
>
>- The Fortran, Java, and C++ MPI examples are now no longer build by
>default. Er... what happened there, and why?
>
>- Why are the oshmem examples in a separate target? The point of the
>previous makefile
Hi,
I've been using nightly tarball openmpi-1.9a1r28919 with Java binding and
it had a bug in its MINLOC implementation (Minloc.java). Essentially, the
following line,
out_array [outdisp + 1] = in_array [*outdisp* + 1] ;
should be changed to
out_array [outdisp + 1] = in_array [*indisp* + 1] ;
On Nov 4, 2013, at 12:14 PM, Sylvestre Ledru wrote:
> FYI, I tried with OpenMPI 1.7.3 under ia64 and the build fails with:
> libtool: link: gcc -std=gnu99 -DNDEBUG -g -O2 -Wformat
> -Werror=format-security -finline-functions -fno-strict-aliasing -pthread
> -o .libs/orte-clean orte-clean.o -L/usr
I'm lost - where were you making a change? I'm unaware of Minloc.java or
Maxloc.java files in the MPI bindings, so I assume the change was being made to
the Intracomm.java file?
On Nov 11, 2013, at 12:35 PM, Saliya Ekanayake wrote:
> Hi,
>
> I've been using nightly tarball openmpi-1.9a1r289
Hi Ralph,
In the nightly snapshot that I was using (openmpi-1.9a1r28919) there were
two java files (Minloc.java and Maxloc.java) inside
$OMPI_HOME/ompi/mpi/java/java
However, these are no longer available in the trunk. I think this is
because of the reason Oscar mentioned that now MINLOC and MAXL
Afraid I'll have to let them answer - I can't find the corresponding code.
On Nov 11, 2013, at 6:50 PM, Saliya Ekanayake wrote:
> Hi Ralph,
>
> In the nightly snapshot that I was using (openmpi-1.9a1r28919) there were two
> java files (Minloc.java and Maxloc.java) inside $OMPI_HOME/ompi/mpi/ja
