On Jun 3, 2010, at 7:34 PM, jtg wrote:

> My FiPy/Trilinos code has a memory leak.  How to recreate the problem:

> While I begin the prayer and fasting rituals required in order to undertake 
> more trilinos builds, I am asking if some kind soul from the list has any 
> insight into this problem or would try this test and report what they see.  
> For example, did I forgot to disable the ENABLE_CATASTROPHIC_MEMORY_GUSHER 
> option in trilinos (Default = True)?

This is probably an issue that Wheeler has just (in the last two days) found 
and fixed in PySparse. Yes, I know that you're only seeing the leak when using 
Trilinos, but we presently require a (faulty) blowout-preventer from PySparse 
when solving with Trilinos. 

I don't think there has been a new release of PySparse quite yet, but the patch 
as been made to both 
http://pysparse.svn.sourceforge.net/viewvc/pysparse/branches/1_1_x/ and 
http://pysparse.svn.sourceforge.net/viewvc/pysparse/trunk/





Reply via email to