On Mon, Apr 08, 2013 at 09:17:48AM -0400, Jordi Gutiérrez Hermoso wrote:
> On 4 April 2013 19:54, Bradley M. Froehle wrote:
> > If only there was a way for the different variants of HDF5 to be
> > simultaneously installable…
>
> I think this is the real problem,and it goes way up to HDF5 upstrea
Hello All,
You should not have to install multiple HDF5 libraries. You can use the
same library for parallel or serial use, which is why it works the way
it does.
Can you give us more information on why you would like to have the
parallel library renamed?
Thanks,
-Barbara
Add HDF5 upstream contact address.
On 08/04/2013 15:17, Jordi Gutiérrez Hermoso wrote:
> On 4 April 2013 19:54, Bradley M. Froehle wrote:
>> If only there was a way for the different variants of HDF5 to be
>> simultaneously installable…
>
> I think this is the real problem,and it goes way up to
On 4 April 2013 19:54, Bradley M. Froehle wrote:
> If only there was a way for the different variants of HDF5 to be
> simultaneously installable…
I think this is the real problem,and it goes way up to HDF5 upstream.
They basically ship two completely different libraries with the same
soname and
> More practically, and more importantly, Octave users who are compiling
> extensions from within Octave are yet another step removed from both the
> compiler and the HDF5 library. They may not even be aware of HDF5 but
> the compilation and linking of their oct-file will be affected.
> Mkoctfile s
On Sun, Mar 31, 2013 at 05:52:03PM -0700, Bradley M. Froehle wrote:
> As an example of how confounding this is, just consider a simple example:
>
> $ cat test.cpp
> #include
> #include
> int main() {
> MPI::Init();
> printf("I'm %d of %d\n", MPI::COMM_WORLD.Get_rank(),
> MPI::COMM_WORLD.Get_
Hi Mike:
Sorry that you'll get this twice. The first time around I accidentally
replied only to you. Please read on for responses inline.
On Sun, Mar 31, 2013 at 4:08 PM, Mike Miller wrote:
> CC'ing pkg-octave-devel since this impacts Octave packaging as well.
Thanks.
> Bradley M. Froehle
CC'ing pkg-octave-devel since this impacts Octave packaging as well.
Bradley M. Froehle wrote:
> The fix for this bug provided in 1.8.9-1~exp2 has cause me a good deal of
> headache today.
>
> For me, the issue is triggered as some C++ code containing:
>
> #include// sets OMPI_SKIP_MPICXX 1
>
I've looked into this bug a bit today and I'd suggest that instead the
`mkoctfile-mpi.diff` patch in src:octave (from bug #598227) be modified to
be something more like:
-: ${XTRA_CXXFLAGS=%OCTAVE_CONF_XTRA_CXXFLAGS%}
+: ${XTRA_CXXFLAGS=-I/usr/include/mpi -DOMPI_SKIP_MPICXX=1
-DMPICH_SKIP_MPICXX=1
The fix for this bug provided in 1.8.9-1~exp2 has cause me a good deal of
headache today.
For me, the issue is triggered as some C++ code containing:
#include// sets OMPI_SKIP_MPICXX 1
#include // MPI C++ namespace now is NOT available
Note that even trying to unset OMPI_SKIP_MPICXX before
On Mon, Sep 10, 2012 at 2:46 PM, Thomas Weber wrote:
> On Fri, Sep 07, 2012 at 09:11:34AM -0400, Mike Miller wrote:
>> mkoctfile -DMPICH_SKIP_MPICXX=1 -DOMPI_SKIP_MPICXX=1 helloworld.cc
>>
>> I don't know enough about HDF5, but if it doesn't make use of the MPI
>> C++ bindings at all, I'd argue thi
On Fri, Sep 07, 2012 at 09:11:34AM -0400, Mike Miller wrote:
> mkoctfile -DMPICH_SKIP_MPICXX=1 -DOMPI_SKIP_MPICXX=1 helloworld.cc
>
> I don't know enough about HDF5, but if it doesn't make use of the MPI
> C++ bindings at all, I'd argue this should be reassigned to hdf5
> upstream so H5public.h wo
reassign 686926 src:hdf5 1.8.8-9
retitle 686926 hdf5: Skip the inclusion of the C++ API
tags 686926 upstream patch
thanks
* Mike Miller [2012-09-07 09:11]:
> On Fri, Sep 7, 2012 at 8:51 AM, Sébastien Villemot wrote:
> > Mike Miller writes:
> >
> >> On Fri, Sep 7, 2012 at 4:50 AM, Sébastien Vill
On Fri, Sep 7, 2012 at 8:51 AM, Sébastien Villemot wrote:
> Mike Miller writes:
>
>> On Fri, Sep 7, 2012 at 4:50 AM, Sébastien Villemot wrote:
>>> Note that this problem is only present with libhdf5-openmpi-dev: it is not
>>> present with libhdf5-dev (serial version, the default) or
>>> libhdf5-m
Mike Miller writes:
> On Fri, Sep 7, 2012 at 4:50 AM, Sébastien Villemot wrote:
>> Note that this problem is only present with libhdf5-openmpi-dev: it is not
>> present with libhdf5-dev (serial version, the default) or libhdf5-mpich2-dev.
>
> Hm, are you sure? I see the same error with either lib
On Fri, Sep 7, 2012 at 4:50 AM, Sébastien Villemot wrote:
> Note that this problem is only present with libhdf5-openmpi-dev: it is not
> present with libhdf5-dev (serial version, the default) or libhdf5-mpich2-dev.
Hm, are you sure? I see the same error with either libhdf5-mpich2-dev
or libhdf5-op
Package: liboctave-dev
Version: 3.6.2-5
Severity: important
When libhdf5-openmpi-dev is co-installed with liboctave-dev, the .oct files
produced by mkoctfile trigger the following error when run from Octave:
error:
/usr/lib/x86_64-linux-gnu/octave/packages/optim-1.2.0/x86_64-pc-linux-gnu-api-v4
17 matches
Mail list logo
