[cig-commits] r12491 - seismo/2D/SPECFEM2D/trunk

nlegoff at geodynamics.org nlegoff at geodynamics.org
Tue Jul 29 16:41:00 PDT 2008


Author: nlegoff
Date: 2008-07-29 16:41:00 -0700 (Tue, 29 Jul 2008)
New Revision: 12491

Modified:
   seismo/2D/SPECFEM2D/trunk/specfem2D.F90
Log:
fixed a bug concerning absorbing condition and Cuthill-McKee : it is now okay to use absorbing conditions along with a Cuthill-McKee reordering of the elements.

Modified: seismo/2D/SPECFEM2D/trunk/specfem2D.F90
===================================================================
--- seismo/2D/SPECFEM2D/trunk/specfem2D.F90	2008-07-29 23:10:24 UTC (rev 12490)
+++ seismo/2D/SPECFEM2D/trunk/specfem2D.F90	2008-07-29 23:41:00 UTC (rev 12491)
@@ -783,7 +783,13 @@
       read(IIN,*) numabsread,codeabsread(1),codeabsread(2),codeabsread(3),codeabsread(4), ibegin_bottom(inum), iend_bottom(inum), &
            jbegin_right(inum), jend_right(inum), ibegin_top(inum), iend_top(inum), jbegin_left(inum), jend_left(inum)
       if(numabsread < 1 .or. numabsread > nspec) call exit_MPI('Wrong absorbing element number')
-      numabs(inum) = numabsread
+      if(ipass == 1) then
+        numabs(inum) = numabsread
+      else if(ipass == 2) then
+        numabs(inum) = perm(antecedent_list(numabsread))
+      else
+        call exit_MPI('error: maximum number of passes is 2')
+      endif
       codeabs(IBOTTOM,inum) = codeabsread(1)
       codeabs(IRIGHT,inum) = codeabsread(2)
       codeabs(ITOP,inum) = codeabsread(3)



More information about the cig-commits mailing list