dolfin team mailing list archive

Thread
Date

[HG DOLFIN] Some clean up of PETScMatrix.

To: dolfin-dev@xxxxxxxxxx
From: dolfin@xxxxxxxxxx (DOLFIN)
Date: Sun, 27 Sep 2009 22:48:14 +0200 (CEST)
Delivered-to: dolfin-dev@xxxxxxxxxx

One or more new changesets pushed to the primary dolfin repository.
A short summary of the last three changesets is included below.

changeset:   7156:12b5cb95c96e
tag:         tip
user:        "Garth N. Wells <gnw20@xxxxxxxxx>"
date:        Sun Sep 27 21:48:03 2009 +0100
files:       dolfin/la/PETScMatrix.cpp dolfin/la/PETScMatrix.h dolfin/swig/docstrings.i
description:
Some clean up of PETScMatrix.


changeset:   7155:04fdd2dbad10
user:        "Garth N. Wells <gnw20@xxxxxxxxx>"
date:        Sun Sep 27 20:11:40 2009 +0100
files:       dolfin/la/PETScFactory.cpp dolfin/la/PETScKrylovMatrix.cpp dolfin/la/PETScKrylovSolver.cpp dolfin/la/PETScKrylovSolver.h dolfin/la/PETScLUSolver.cpp dolfin/la/PETScMatrix.cpp dolfin/la/PETScMatrix.h dolfin/la/PETScVector.cpp dolfin/la/SLEPcEigenSolver.cpp
description:
Remove support for PETSc 2.3.3. DOLFIN now requires PETSc 3.0.0.

The next step is to clean up the logic in some of the PETSc wrappers. Logic was strange since to use a particular solver, the a matrix had to be set to that 'type'. In PETSc 3 the solver type is set in the solver (more sensible).


changeset:   7154:c0192852f3d8
user:        "Garth N. Wells <gnw20@xxxxxxxxx>"
date:        Sun Sep 27 11:45:04 2009 +0100
files:       dolfin/fem/SystemAssembler.cpp test/unit/fem/python/test.py
description:
Small bug fix in SystemAssembler.

----------------------------------------------------------------------
For more details, visit http://www.fenics.org/hg/dolfin