if multiplicative mode uses multiple cycles per apply in the preconditioner then initial guess for down smoother will be nonzero and hence needs to be marked

Reported-by: Alirezaa Jalaali <alirezaa.

if multiplicative mode uses multiple cycles per apply in the preconditioner then initial guess for down smoother will be nonzero and hence needs to be marked

Reported-by: Alirezaa Jalaali <alirezaa.jalaali@gmail.com>

show more ...

Merge branch 'maint'

PCMG must set initial guess nonzero for down (pre) smoother when using full or Kaskade multigrid because otherwise it zeros out the interpolated solution from the coarser level

Reported-by: Alirezaa

PCMG must set initial guess nonzero for down (pre) smoother when using full or Kaskade multigrid because otherwise it zeros out the interpolated solution from the coarser level

Reported-by: Alirezaa Jalaali <alirezaa.jalaali@gmail.com>

show more ...

fixes and improvements to a variety of manual pages

improvements to GAMG documentation including cleanup of manual pages, more links between manual pages and links in the users manual to all the manual pages

Merge branch 'tisaac/feature-gamg-extensible' into barry/experiment

Conflicts:
src/ksp/pc/impls/gamg/gamg.h

Move MG and GAMG implementation headers to petsc-private

I want to be able to register GAMG plugins. I considered a few options:

a) Move the implementation headers to petsc-private so they can be

Move MG and GAMG implementation headers to petsc-private

I want to be able to register GAMG plugins. I considered a few options:

a) Move the implementation headers to petsc-private so they can be
installed

b) Make Get/Set methods for all of the GAMGOps: graph, coarsen,
optprol, etc.

c) Add a PCGAMGGet/SetOperation like MatShellGet/SetOperation for each
of these operations.

The easiest was (a), and the GAMG methods are so context-specific I'm
not sure who would benefit from their being part of the public petsc
interface. If GAMG get's refactored to where the operations are easier
to understand on their own, then maybe (b) or (c) would be a better
option.

show more ...

simplify option object names

moved PetscOptionsObject from global variable to local variable in functions

manual merge of conflict

Merge branch 'maint'

allow PCMG user to provide either interpolation or restriction operator with Galerkin coarse grid operators

Merge branch 'barry/fix-setters'

Merge branch 'maint'

Conflicts:
config/BuildSystem/config/packages/ml.py
src/ksp/pc/impls/mg/mg.c

fix formatting of some manual pages
specifically remove use of $ from the middle of parameter or options lists

added a variety of getters and fixed calling sequence of a few setters

Merge branch 'barry/fix-options-default'

Conflicts:
src/ksp/ksp/examples/tutorials/ex52.c

cleanup of use of PetscOptionsInt and friends including use of flag final argument

MatGetVecs() and KSPGetVecs() replaced with MatCreateVecs() and KSPCreateVecs()
since they are creators not getters

MG options documentation fix