[petsc-dev] Fortran seems broken

Satish Balay balay at mcs.anl.gov
Fri Jan 23 14:37:04 CST 2015


On Fri, 23 Jan 2015, Mark Adams wrote:

> On Fri, Jan 23, 2015 at 3:00 PM, Satish Balay <balay at mcs.anl.gov> wrote:
> 
> > The example with errors - is using: PETSC_ARCH=arch-xe6-opt64-pgi
> > configure.log is using: PETSC_ARCH=arch-xe6-opt64-gnu
> > make.log sent is using: PETSC_ARCH=arch-xc30-opt64-intel
> >
> >
> I was doing multiple configures at the same time.  Should you be able to do
> that?

Yes you can do multiple simultaneous builds [with some caveats] - but
when sending logs - you should send the relevant ones for the case you
have errors.

In your case (due to multiple simultaneous builds) - the logs in
PETSC_DIR might not be the relavent ones - so you could have send the
ones from the relavent PETSC_DIR/PETSC_ARCH/lib/petsc-conf/ location.

Satish

> 
> 
> > Can you send the logs [and ls] for the correct build you need to use?
> > And are you sure you are using the appropriate compilers -
> > correspoding to the build?
> >
> 
> I am starting from scratch now.
> Thanks,
> 
> 
> >
> > Satish
> >
> > On Fri, 23 Jan 2015, Mark Adams wrote:
> >
> > > On Fri, Jan 23, 2015 at 2:42 PM, Satish Balay <balay at mcs.anl.gov> wrote:
> > >
> > > > What do you have for:
> > > >
> > > > ls -l /global/u2/m/madams/petsc_private/arch-xc30-opt64-intel/include/
> > > >
> > > >
> > > 11:51 edison01 master ~/petsc_private/arch-xc30-opt64-intel$ ls -l
> > > /global/u2/m/madams/petsc_private/arch-xc30-opt64-intel/include/
> > > total 620
> > > -rw-r----- 1 madams madams 13945 Jan 23 07:13 Cnames.h
> > > -rw-r----- 1 madams madams  1749 Jan 23 07:13 dcomplex.h
> > > -rw-r--r-- 1 madams madams  1631 Mar 26  2012 gklib_defs.h
> > > -rw-r--r-- 1 madams madams  4018 Mar 26  2012 gklib_rename.h
> > > -rw-r--r-- 1 madams madams    60 Jan 23 06:58 gklib_tls.h
> > > -rw-r----- 1 madams madams   433 Jan 23 07:13 html_mainpage.h
> > > -rw-r----- 1 madams madams   914 Jan 23 07:13 machines.h
> > > -rw-r--r-- 1 madams madams 10355 Jan 23 06:58 metis.h
> > > -rw-r----- 1 madams madams  2610 Jan 23 07:13 old_colamd.h
> > > -rw-r--r-- 1 madams madams  6121 Aug 21  2013 parmetis.h
> > > -rw-r----- 1 madams madams 11021 Jan 23 07:22 petsc.mod
> > > -rw-r----- 1 madams madams 10928 Jan 23 07:22 petscao.mod
> > > -rw-r----- 1 madams madams 11382 Jan 23 07:22 petscaodef.mod
> > > -rw-r----- 1 madams madams 15518 Jan 23 07:19 petscconf.h
> > > -rw-r----- 1 madams madams   555 Jan 23 07:17 petscconfiginfo.h
> > > -rw-r----- 1 madams madams 10933 Jan 23 07:22 petscdef.mod
> > > -rw-r----- 1 madams madams 10927 Jan 23 07:22 petscdm.mod
> > > -rw-r----- 1 madams madams 17751 Jan 23 07:22 petscdmcomposite.mod
> > > -rw-r----- 1 madams madams 22412 Jan 23 07:22 petscdmda.mod
> > > -rw-r----- 1 madams madams 12229 Jan 23 07:22 petscdmdadef.mod
> > > -rw-r----- 1 madams madams 11702 Jan 23 07:22 petscdmdef.mod
> > > -rw-r----- 1 madams madams   123 Jan 23 07:19 petscfix.h
> > > -rw-r----- 1 madams madams 18109 Jan 23 07:22 petscis.mod
> > > -rw-r----- 1 madams madams 11965 Jan 23 07:22 petscisdef.mod
> > > -rw-r----- 1 madams madams 10931 Jan 23 07:22 petscksp.mod
> > > -rw-r----- 1 madams madams 15980 Jan 23 07:22 petsckspdef.mod
> > > -rw-r----- 1 madams madams  1779 Jan 23 07:17 petscmachineinfo.h
> > > -rw-r----- 1 madams madams 30687 Jan 23 07:22 petscmat.mod
> > > -rw-r----- 1 madams madams 47664 Jan 23 07:22 petscmatdef.mod
> > > -rw-r----- 1 madams madams 10930 Jan 23 07:22 petscpc.mod
> > > -rw-r----- 1 madams madams 18231 Jan 23 07:22 petscpcdef.mod
> > > -rw-r----- 1 madams madams 10936 Jan 23 07:22 petscsnes.mod
> > > -rw-r----- 1 madams madams 17890 Jan 23 07:22 petscsnesdef.mod
> > > -rw-r----- 1 madams madams 15311 Jan 23 07:19 petscsys.mod
> > > -rw-r----- 1 madams madams 35642 Jan 23 07:19 petscsysdef.mod
> > > -rw-r----- 1 madams madams 10929 Jan 23 07:22 petscts.mod
> > > -rw-r----- 1 madams madams 15473 Jan 23 07:22 petsctsdef.mod
> > > -rw-r----- 1 madams madams 17328 Jan 23 07:22 petscvec.mod
> > > -rw-r----- 1 madams madams 14292 Jan 23 07:22 petscvecdef.mod
> > > -rwxr-x--- 1 madams madams 12019 Jan 23 07:13 psymbfact.h
> > > -rw-r----- 1 madams madams 15133 Jan 23 07:13 superlu_ddefs.h
> > > -rw-r----- 1 madams madams 25138 Jan 23 07:13 superlu_defs.h
> > > -rw-r----- 1 madams madams  2984 Jan 23 07:13 superlu_enum_consts.h
> > > -rw-r----- 1 madams madams 15920 Jan 23 07:13 superlu_zdefs.h
> > > -rw-r----- 1 madams madams  7722 Jan 23 07:13 supermatrix.h
> > > -rw-r----- 1 madams madams  3506 Jan 23 07:13 util_dist.h
> > > 11:51 edison01 master ~/petsc_private/arch-xc30-opt64-intel$
> > >
> > >
> > > > Satish
> > > >
> > > > On Fri, 23 Jan 2015, Mark Adams wrote:
> > > >
> > > > > oh and I do have the logs.
> > > > >
> > > > > On Fri, Jan 23, 2015 at 12:56 PM, Matthew Knepley <knepley at gmail.com
> > >
> > > > wrote:
> > > > >
> > > > > > It wants the petscmat.mod module. Can you send the compile line?
> > > > > >
> > > > > >    Matt
> > > > > >
> > > > > > On Fri, Jan 23, 2015 at 11:54 AM, Mark Adams <mfadams at lbl.gov>
> > wrote:
> > > > > >
> > > > > >> This code has started failing.  I've change finclude
> > -->petsc-finclude
> > > > > >>
> > > > > >> #include <petsc-finclude/petscdef.h>
> > > > > >>   use petscmat
> > > > > >>
> > > > > >> I get:
> > > > > >>
> > > > > >> fem_ops.F90(19): error #7002: Error in opening the compiled module
> > > > file.
> > > > > >> Check INCLUDE paths.   [PETSCMAT]
> > > > > >>   use petscmat
> > > > > >>
> > > > > >> I've clobbered the confugure.log but can regenerate it if need be.
> > > > > >>
> > > > > >> Thanks,
> > > > > >> Mark
> > > > > >>
> > > > > >>
> > > > > >
> > > > > >
> > > > > > --
> > > > > > 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
> > > > > >
> > > > >
> > > >
> > > >
> > >
> >
> >
> 




More information about the petsc-dev mailing list