Added support for finding out how many surfaces are present and then do the

integrals over all surfaces. Still only supports full mode. Only weakly tested.
There is a problem with the outputs in multiprocessor mode. The results of
the integration are stored in 1D grid arrays. However if I use DISTRIB=DEFAULT
(taking care only to store results on the correct processor), the
output fails if the number of processors are bigger than the number of
elements in the grid array (essentially some processors contains a chunk
of length zero). If I on the other hand uses DISTRIB=CONSTANT, it works
but on multiple processors i get multiple copies of the data in the output.


git-svn-id: http://svn.einsteintoolkit.org/cactus/EinsteinAnalysis/EHFinder/trunk@90 2a26948c-0e4f-0410-aee8-f1d3e353619c

1 files changed, 2 insertions, 1 deletions

diff --git a/param.ccl b/param.ccl
index ca7335f..b48dd45 100644
--- a/param.ccl
+++ b/param.ccl
@@ -178,7 +178,8 @@ INT n_array_ghosts "Number of ghost points in the surface grid array"
 1:               :: "Positive please"
 } 1
 
-INT maximum_surface_number "The maximum number of surfaces expected in the data"{
+INT maximum_surface_number "The maximum number of surfaces expected in the data"
+{
   1:*            :: "Positive please"
 } 1