On Mon, Sep 1, 2014 at 6:57 AM, Åsmund Ervik <[email protected]> wrote:
> > On 01. sep. 2014 13:52, Matthew Knepley wrote: > > On Mon, Sep 1, 2014 at 6:18 AM, Åsmund Ervik <[email protected]> > wrote: > > > >> > >> > >> On 01. sep. 2014 12:51, Matthew Knepley wrote: > >>> On Mon, Sep 1, 2014 at 3:22 AM, Åsmund Ervik <[email protected]> > >> wrote: > >>> > >>>> > >>>> Subsequently I installed HDF5 v. 1.8.12 using my OS package manager > and > >>>> then tried to configure with "--with-hdf5 > >>>> --with-hdf5-dir=/path/to/system/hdf5" using PETSc 3.5.1. This time > both > >>>> configure and make were successful, but "make test" fails with loads > of > >>>> undefined references to PETSc HDF5 stuff (see below). configure.log > and > >>>> make.log attached also for this case (error-with-hdf5.tar.gz) > >>>> > >>> > >>> It did not build src/sys/classes/viewer/impls/hdf5/, clearly from the > >> log. > >>> However, it > >>> should have. We have been trying to understand why Make behaves in a > bad > >> way > >>> after configuration failure. It should go away with make clean and > >> another > >>> make. > >> > >> Thanks Matt, make clean and then make did result in make test passing > >> all tests. > >> > >> However, I am still unable to use HDF5 functionality. When I compile and > >> run e.g. vec/vec/examples/tutorials/ex10.c it works fine with binary > >> output, but with HDF5 I get the error message below. > >> > >> I'm guessing this is some incompatibility between the various MPI, HDF5 > >> and PETSc versions, i.e. that the HDF5 from my OS is not using the same > >> MPI as PETSc. Is it simple to edit something in PETSc and have > >> "./configure --download-hdf5" get the most recent HDF5 library which > >> compiles on my machine? > >> > > > > Yes, I cannot reproduce, so it must be something like that. Can you > > reconfigure > > using the --download version? > > No, I am not able to, that is why I switched to "--with-hdf5-dir=". Like > I said in the first email, the version of HDF5 (1.8.10) that ships with > PETSc 3.5.1 does not compile on my machine. This is apparently a known > bug that was fixed upstream, cf. > > https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=711777 > > This is the reason why I asked whether I can somehow tell > "--download-hdf5" to download a more recent version. 1.8.11 should do it. People who leave commented out code in there should be tarred, feathered, and run out of town on a rail. You can just go in and delete that line, and then reconfigure. It will use the source that is already downloaded. Satish, can we patch our download source? I can do it, but I am not sure of the process for HDF5. Thanks, Matt > > Åsmund > > -- What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead. -- Norbert Wiener
