[petsc-dev] RDict import fragility

[Matthew Knepley]

On Mon, Aug 5, 2013 at 8:54 AM, Jed Brown <jedbrown at mcs.anl.gov> wrote:

> When I try to configure SLEPc, I get
>
>   ERROR: Importing cmakeboot failed:
>   No module named txpetscgpu
>

I guess we could prune modules that are not found before saving. I was
under the impression
that reconfiguring after source changes to BuildSystem was not a big deal.

   Matt


> Traceback (most recent call last):
>   File "./configure", line 10, in <module>
>     execfile(os.path.join(os.path.dirname(__file__), 'config',
> 'configure.py'))
>   File "./config/configure.py", line 403, in <module>
>     cmakeok =
> cmakeboot.main(slepcdir,petscdir,petscarch=petscconf.ARCH,log=log)
>   File "/home/jed/slepc/config/cmakeboot.py", line 165, in main
>     return
> PETScMaker(slepcdir,petscdir,petscarch,argDB,framework).cmakeboot(args,log)
>   File "/home/jed/slepc/config/cmakeboot.py", line 80, in cmakeboot
>     self.setup()
>   File "/home/jed/slepc/config/cmakeboot.py", line 75, in setup
>     self.framework = self.loadConfigure()
>   File "/home/jed/petsc/config/BuildSystem/script.py", line 271, in
> loadConfigure
>     framework = cPickle.loads(cache)
>
> It appears that if a module is removed from PETSc, even if it's not
> used, the import will fail.  That's really annoying because it forces a
> reconfigure even for all the PETSC_ARCHes that never used txpetscgpu.
>



-- 
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
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-dev/attachments/20130805/5321c3c7/attachment.html>