dolfin team mailing list archive

[Johan Hake <702372@xxxxxxxxxxxxxxxxxx>]

On Thursday January 13 2011 23:41:26 Joachim Haga wrote:
> I understand there is a larger issue here. In the meantime, Johan's fix
> (which just keeps the submatrices referenced on the python level) is
> sufficient to remove the crash. Or mine, but Johan's has the distinct
> advantage of being already committed :)

I guess you are right! Sorry for that ;)
 
> Be aware that there is a memory leak in Block* if "owner" is set on the
> constructor and "set(...)" is later called, but that's probably not very
> important.

Ok, I think the best fix is the one that Garth started out on, using 
shared_ptrs. But there are some issues as all objects inheriting from 
GenericTensor then needs to be stored using shared_ptr in Python. This mostly 
works out of the box, but we probably need to construct director typemaps or 
add some shared_ptr<GenericFoo> arguments to director classes, as SWIG wont 
handle these correctly.

On the TODO list!

Johan

-- 
You received this bug notification because you are a member of DOLFIN
Team, which is subscribed to DOLFIN.
https://bugs.launchpad.net/bugs/702372

Title:
  BlockMatrix and deleted matrices

Status in DOLFIN:
  New

Bug description:
  The following code crashes, unless "del A" is commented out. I don't
  know the memory handling strategy in dolfin, (boost::shared_ptr is
  used in some places, but not consistently?) but attached is a patch
  which at least removes the crash by making private copies of the
  matrices (and vectors in BlockVectors).

  from dolfin import *

  mesh = UnitSquare(4, 4)
  V = FunctionSpace(mesh, "CG", 1)
  u,v = TrialFunction(V), TestFunction(V)

  a = u*v*dx

  AA = BlockMatrix(1,1)

  A = assemble(a)
  AA[0, 0] = A

  del A

  print AA[0,0]