add parameter to disable MoL_PostStep in Post_Recover_Variables

this is to allow bit-wise recovery of Hydro data since con2prim runs in MoL_PostStep. Also we should be able to read in all data from a checkpoint and thus mol_poststep should not be required. Thorns that want to be clever and re-compute some quantities that can be recomputed from the checkpointed data should schedule themselves explicitly in Post_Recover_Variables.

See extensive discussion at https://trac.einsteintoolkit.org/ticket/1256


git-svn-id: http://svn.cactuscode.org/arrangements/CactusNumerical/MoL/trunk@199 578cdeb0-5ea1-4b81-8215-5a3b8777ee0b

2 files changed, 7 insertions, 0 deletions

diff --git a/param.ccl b/param.ccl
index 34282d1..a6c973c 100644
--- a/param.ccl
+++ b/param.ccl
@@ -167,6 +167,11 @@ BOOLEAN copy_ID_after_MoL_PostStep \
 {
 } "no"
 
+BOOLEAN run_MoL_PostStep_in_Post_Recover_Variables "Schedule the PostStep parts after recovery so that symmetries are automatically done correctly." STEERABLE=recover
+{
+} "yes"
+
+
 BOOLEAN set_ID_boundaries "Should boundaries be overwritten (via synchronization, prolongation, boundary conditions) by MoL?"
 {
 } "yes"
diff --git a/schedule.ccl b/schedule.ccl
index 6d607ba..44f9b6d 100644
--- a/schedule.ccl
+++ b/schedule.ccl
@@ -706,9 +706,11 @@ schedule GROUP MoL_PostStep AT PostRestrict
 ###  expensive, but it's simplest for the moment.          ###
 ##############################################################
 
+if(run_MoL_PostStep_in_Post_Recover_Variables) {
 schedule GROUP MoL_PostStep AT Post_Recover_Variables
 {
 } "Ensure that everything is correct after recovery"
+}
 
 #########################################################################
 ###  Physics thorns can enforce constraints in PostStepModify.        ###