[Solved] Mesh Update in ResultOutputSolve

[Martina]

Hi,

I am encountering problems with saving Mesh Update with ResultOuputSolve with r6262 which did not occur with r6235.

Using ResultOutputSolve in a run that does not involve the Mesh Update Solver without indicating the list of Scalars and Vectors to be saved works. Indicating a list which contains Mesh Update also works. However not indicating a list and using Mesh Update Solver leads to a Seg.fault.

Code: Select all

ResultOutputSolver: -------------------------------------
ResultOutputSolve: Saving with prefix: pv.asf_ti_0.sif.debug.
ResultOutputSolver: Working on mesh: mesh
ResultOutputSolver: Dimension of mesh: 3
ResultOutputSolver: Creating list for saving - if not present
reateListForSaving: Field Variables for Saving
CreateListForSaving: Scalar Field 1: pressure
CreateListForSaving: Scalar Field 2: beta
XXXXX
CreateListForSaving: Vector Field 1:
CreateListForSaving: Vector Field 2: velocity

In the automatically created List of variables appears a Vector field without name. I'm quite sure that this is supposed to be Mesh Update since I do not have any other vector fields (also ElmerPost shows the Vector fields Mesh Update and Velocity).

Later I get the warnings

Code: Select all

WARNING:: WriteVtuXMLFile: Nonexistent variable 1:
WARNING:: WriteVtuXMLFile: Nonexistent variable 1:
WARNING:: WriteVtuXMLFile: Nonexistent variable 3:

and finally a segmentation fault because of these Nonexisting variables.

Martina

[raback]

Sorry for that, it should be fixed now. Thanx again for the diligent error report. -Peter

[Martina]

Great thanks, it works.

A related question: is there a reason why

Code: Select all

WARNING:: WriteVtuXMLFile: Nonexistent variable

warnings lead to a seg.fault? Couldn't it just remain warning (or for consistency be an error)? I find it often quite annoying having a simulation run (correctly) over hours and then to end in a seg.fault only because of a typing error in the Variable list of ResultOutputSolve (or having by error added a Variable which shouldn't be there). It would be really great if at least the other variables could be written and the simulation go on...

Martina

[raback]

Hi Martina

Yes, you're right. There happened to be one Fatal left. I replaced with Warn & CYCLE. So now not everything will be lost.

-Peter