dolfin team mailing list archive

["Garth N. Wells" <gnw20@xxxxxxxxx>]

Johan Hake wrote:
> On Friday 02 January 2009 09:53:06 Johan Hake wrote:
> > On Friday 02 January 2009 09:27:13 Anders Logg wrote:
> > > The buildbots look green and most things look ok from here.
> > >
> > > I suggest we make the release of 0.9.0 on Monday Jan 5. Shout if there
> > > is anything else we need to wait for (or fix it before Monday).
> > >
> > > There is always 0.9.1, 0.9.2 etc so everything does not need to be in
> > > this release. We've made big interface changes so it's good to get out
> > > a new "stable" release for use and testing.
> > The simple test script below fails on me. MatDuplicate() complains of
> >
> >   Null argument, when expecting valid pointer!
> >   Null Pointer: Parameter # 3!
> >
> > This did not happen before the change to shared_ptr for the Mat. Can any
> > with more PETSc wiz look into this?
> >
> > Johan
>
> Btw, this error is visable when the python part of test.py in test/unit/ is 
> uncommented. 
>
> We should uncomment these before a release. Now it is only the error above 
> that fails on my computer.

I've turned the Python tests back on, and the memory check. Everything 
passes on my machine.

It would be good to undo the change to PETSc on the buildbot (enabling 
C++) to make sure that we're not using any C++ calls to PETSc.

Garth


> Johan
>
>
> > *************************************************
> >
> > from dolfin import *
> >
> > dolfin_set("linear algebra backend","PETSc")
> >
> > m = UnitSquare(100,100)
> > V = FunctionSpace(m,"Lagrange",1)
> >
> > u = TrialFunction(V)
> > v = TestFunction(V)
> >
> > A = assemble(u*v*dx)
> >
> > a = A(1,1)
> > A2 = A.copy()
> > _______________________________________________
> > DOLFIN-dev mailing list
> > DOLFIN-dev@xxxxxxxxxx
> > http://www.fenics.org/mailman/listinfo/dolfin-dev
>
>
> _______________________________________________
> DOLFIN-dev mailing list
> DOLFIN-dev@xxxxxxxxxx
> http://www.fenics.org/mailman/listinfo/dolfin-dev