normalize does not work in parallel

[Mikael Mortensen]

This code seems to fail in parallel because the vector has different local size on each processor. (The code works for UnitSquare(2,2)):

from dolfin import *
mesh = UnitSquare(5, 5)
V = FunctionSpace(mesh, 'CG', 1)
u = Function(V)
u.vector()[:] = 1.
normalize(u.vector())

A fix is committed in the attached patch.

[Mikael Mortensen (mikael-mortensen)]

Sorry, that should be local_range in the patch, not local_size. Correct
patch is attached here.

M

On 6 December 2011 15:14, Mikael Mortensen <email address hidden>wrote:

> Public bug reported:
>
> This code seems to fail in parallel because the vector has different
> local size on each processor. (The code works for UnitSquare(2,2)):
>
> from dolfin import *
> mesh = UnitSquare(5, 5)
> V = FunctionSpace(mesh, 'CG', 1)
> u = Function(V)
> u.vector()[:] = 1.
> normalize(u.vector())
>
> A fix is committed in the attached patch.
>
> ** Affects: dolfin
> Importance: Undecided
> Status: New
>
> --
> You received this bug notification because you are a member of DOLFIN
> Team, which is subscribed to DOLFIN.
> https://bugs.launchpad.net/bugs/900765
>
> Title:
> normalize does not work in parallel
>
> Status in DOLFIN:
> New
>
> Bug description:
> This code seems to fail in parallel because the vector has different
> local size on each processor. (The code works for UnitSquare(2,2)):
>
> from dolfin import *
> mesh = UnitSquare(5, 5)
> V = FunctionSpace(mesh, 'CG', 1)
> u = Function(V)
> u.vector()[:] = 1.
> normalize(u.vector())
>
> A fix is committed in the attached patch.
>
> To manage notifications about this bug go to:
> https://bugs.launchpad.net/dolfin/+bug/900765/+subscriptions
>