On Sat, 24 Aug 2013, subramanya sadasiva wrote:

> In my multi physics solver, I have a couple of explicit equation
> ssystems to store stress and some other processed results. I compute
> the fields and just set them into the vector by going over the
> nodes. However, the code seems to be over-writing one of the fields.
> I checked the system number and the variable number that I am
> setting the variables into  and everything seems okay. is there
> anything else that I could check. Thanks, Subramanya

If this is in parallel, are you sure you're setting values on the
processors that own each node?  Are you synchronizing ghost nodes
before trying to read them on processors that don't own them?

In any case: is it hanging nodes where you see the problem?

Where are the values getting changed?  You might step through with
gdb.
---
Roy

------------------------------------------------------------------------------
Introducing Performance Central, a new site from SourceForge and 
AppDynamics. Performance Central is your source for news, insights, 
analysis and resources for efficient Application Performance Management. 
Visit us today!
http://pubads.g.doubleclick.net/gampad/clk?id=48897511&iu=/4140/ostg.clktrk
_______________________________________________
Libmesh-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/libmesh-users

Reply via email to