On Wed, Feb 23, 2011 at 5:17 PM, Satish Balay <[email protected]> wrote:

It appers to compile only a bit of fortran modules - and not the c sources for me...

Yes, I am fixing that. It is an initialization problem. Since my database is already initialized, I did not have it. Should work soon.

And it leaves arround *.d files in my PETSC_DIR

Yes, that way I can rebuild dependencies at any time. Matt

Satish

-------

Building dependencies in /home/balay/tmp/petsc-dev/include Building dependencies in /home/balay/tmp/petsc-dev/include/sieve Building dependencies in /home/balay/tmp/petsc-dev/include/finclude Building dependencies in /home/balay/tmp/petsc-dev/include/finclude/ftn-auto Building dependencies in /home/balay/tmp/petsc-dev/include/finclude/ftn-custom Building dependencies in /home/balay/tmp/petsc-dev/include/private Building /home/balay/tmp/petsc-dev/src/sys/f90-mod/petscsysmod.F Building /home/balay/tmp/petsc-dev/src/vec/f90-mod/petscvecmod.F Building /home/balay/tmp/petsc-dev/src/mat/f90-mod/petscmatmod.F Building /home/balay/tmp/petsc-dev/src/dm/f90-mod/petscdmmod.F Building /home/balay/tmp/petsc-dev/src/ksp/f90-mod/petsckspmod.F Building /home/balay/tmp/petsc-dev/src/snes/f90-mod/petscsnesmod.F Building /home/balay/tmp/petsc-dev/src/ts/f90-mod/petsctsmod.F asterix:/home/balay/tmp/petsc-dev>ls asterix64/ config/ include/ petsckspmod.d petsctsmod.d setup.py* xcode/ bin/ configure* makefile petscmatmod.d petscvecmod.d src/ CMakeLists.txt configure.log@ make.log@ petscsnesmod.d projects/ TAGS conf/ CTAGS petscdmmod.d petscsysmod.d RDict.log tutorials/ asterix:/home/balay/tmp/petsc-dev>ls asterix64/lib/ libpetsc.a libpetsc-obj/ asterix:/home/balay/tmp/petsc-dev>ar t asterix64/lib/libpetsc.a petscsysmod.o petscvecmod.o petscmatmod.o petscdmmod.o petsckspmod.o petscsnesmod.o petsctsmod.o asterix:/home/balay/tmp/petsc-dev>

On Wed, 23 Feb 2011, Matthew Knepley wrote:

...

I have pushed a new version:

1) F90 modules now build correctly

2) It has been refactored so that all dependency analysis is done up front, and all building happens afterwards.

3) It could take other dependency trackers (insert your favorite)

4) It should be ready for a nightly build

5) It should now be possible to use subprocess to completely parallelize the build.

Alert me any problems that come up. It should work for Windows as well, and be much faster.

Matt

-- 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