Re: [petsc-dev] compiling PETSc Fortran program with modules

2014-06-05 Thread Adrian Croucher
On 05/06/14 23:14, Matthew Knepley wrote: On Thu, Jun 5, 2014 at 12:56 AM, Adrian Croucher mailto:a.crouc...@auckland.ac.nz>> wrote: OK, I found one way around this- if in test.F90 I replace Can you try just using the finclude/petscdef.h90 headers in the module? Aha, yes that does work

Re: [petsc-dev] compiling PETSc Fortran program with modules

2014-06-05 Thread Matthew Knepley
On Thu, Jun 5, 2014 at 12:56 AM, Adrian Croucher wrote: > OK, I found one way around this- if in test.F90 I replace > Can you try just using the finclude/petscdef.h90 headers in the module? Thanks, Matt > use my_module > > with > > use my_module, only : foo > > it works. > > I'm n

Re: [petsc-dev] compiling PETSc Fortran program with modules

2014-06-04 Thread Adrian Croucher
OK, I found one way around this- if in test.F90 I replace use my_module with use my_module, only : foo it works. I'm not sure how practical that's going to be with real code though (e.g. a big module and I want to use lots of things in it). Are there any better ways? - Adrian -- Dr A

[petsc-dev] compiling PETSc Fortran program with modules

2014-06-04 Thread Adrian Croucher
hi I'm having trouble compiling a PETSc Fortran program, which uses a Fortran module in a separate file, which itself uses PETSc. Here's a fairly minimal example of the problem. Main program test.F90: --- program test use my_module implicit none #include PetscInt :: bar = 2 print *