On Wed, Feb 23, 2011 at 5:17 PM, Satish Balay <span dir="ltr"><<a href="mailto:balay@mcs.anl.gov">balay@mcs.anl.gov</a>></span> wrote:<br><div class="gmail_quote"><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
It appers to compile only a bit of fortran modules - and not the c sources for me...<br></blockquote><div><br></div><div>Yes, I am fixing that. It is an initialization problem. Since my database is already initialized,</div>
<div>I did not have it. Should work soon.</div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">
And it leaves arround *.d files in my PETSC_DIR<br></blockquote><div><br></div><div>Yes, that way I can rebuild dependencies at any time.</div><div><br></div><div>   Matt</div><div> </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex;">

Satish<br>
<br>
-------<br>
<br>
Building dependencies in /home/balay/tmp/petsc-dev/include<br>
Building dependencies in /home/balay/tmp/petsc-dev/include/sieve<br>
Building dependencies in /home/balay/tmp/petsc-dev/include/finclude<br>
Building dependencies in /home/balay/tmp/petsc-dev/include/finclude/ftn-auto<br>
Building dependencies in /home/balay/tmp/petsc-dev/include/finclude/ftn-custom<br>
Building dependencies in /home/balay/tmp/petsc-dev/include/private<br>
Building /home/balay/tmp/petsc-dev/src/sys/f90-mod/petscsysmod.F<br>
Building /home/balay/tmp/petsc-dev/src/vec/f90-mod/petscvecmod.F<br>
Building /home/balay/tmp/petsc-dev/src/mat/f90-mod/petscmatmod.F<br>
Building /home/balay/tmp/petsc-dev/src/dm/f90-mod/petscdmmod.F<br>
Building /home/balay/tmp/petsc-dev/src/ksp/f90-mod/petsckspmod.F<br>
Building /home/balay/tmp/petsc-dev/src/snes/f90-mod/petscsnesmod.F<br>
Building /home/balay/tmp/petsc-dev/src/ts/f90-mod/petsctsmod.F<br>
asterix:/home/balay/tmp/petsc-dev>ls<br>
asterix64/      config/         include/      petsckspmod.d   petsctsmod.d   setup.py*   xcode/<br>
bin/            configure*      makefile      petscmatmod.d   petscvecmod.d  src/<br>
CMakeLists.txt  configure.log@  make.log@     petscsnesmod.d  projects/      TAGS<br>
conf/           CTAGS           petscdmmod.d  petscsysmod.d   RDict.log      tutorials/<br>
asterix:/home/balay/tmp/petsc-dev>ls asterix64/lib/<br>
libpetsc.a  libpetsc-obj/<br>
asterix:/home/balay/tmp/petsc-dev>ar t asterix64/lib/libpetsc.a<br>
petscsysmod.o<br>
petscvecmod.o<br>
petscmatmod.o<br>
petscdmmod.o<br>
petsckspmod.o<br>
petscsnesmod.o<br>
petsctsmod.o<br>
asterix:/home/balay/tmp/petsc-dev><br>
<div><div></div><div class="h5"><br>
<br>
On Wed, 23 Feb 2011, Matthew Knepley wrote:<br>
<br>
> I have pushed a new version:<br>
><br>
>   1) F90 modules now build correctly<br>
><br>
>   2) It has been refactored so that all dependency analysis is done up<br>
> front, and<br>
>       all building happens afterwards.<br>
><br>
>   3) It could take other dependency trackers (insert your favorite)<br>
><br>
>   4) It should be ready for a nightly build<br>
><br>
>   5) It should now be possible to use subprocess to completely parallelize<br>
> the build.<br>
><br>
> Alert me any problems that come up. It should work for Windows as well, and<br>
> be<br>
> much faster.<br>
><br>
>    Matt<br>
><br>
><br>
<br>
</div></div></blockquote></div><br><br clear="all"><br>-- <br>What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead.<br>
-- Norbert Wiener<br>