There is a problem compiling upstream
Hierarchy::recalculateGalerkin() without MPI. Therefore we
reimplement that method in our AMG version. Fixes will travel
upstream.
Apart from that we made sure that all types have a meaning even
if there is no MPI found and that we at least have redistribute method
to call even if using SequentialInformation. In that case the method
does nothing but prevents compilation errors complaining about missing
parallel functionality.
As the AMG hierarchy is kept and uses the initial OwnerOverlopCommunication object
we need to make sure that this object lives as long as the hierarchy. We assure
this by storing a shared_ptr to it in ISTLSolverEbosCpr and reuse it to construct
the WellModelMatrixAdapter.
Previously the OwnerOverlapCommunication object was owned by the adapter as a
unique_ptr was used. The change to shared_ptr makes it possible to to extract
it and reuse even if OwnerOverlapCommunication is thrown away.
Somehow some standard output when running MILU () is still there and
it really clutters the output and and makes it unusable. This commit
removes it from flow.
This seems to be covered for types and functions by our coding style
with room for interpretation. For variables the coding styles asks for
underlines though, but nevermind.
The former order resulted of first apply NCC to the grid
transmissibilities and then applying EDITNNC resulted in NNCs being
scaled twices. The reason is that applyNNCToGridTrans_ scales the NNC
with EDITNNC. With the patch the order of the function calls is
reversed to prevent double scaling.
This is includes neighboring connection and NNCs due to faults. In both
cases the transmissibilities of specified via NNC are added to the set or
computed ones.
This is the first step for supporting NNC in flow.
Previously, this method did not take parallelism into account but just
checked the local wells for the constraints. Depending on the load balancing
of the wells this sometimes led to different return values of the function on different
processors. As the output is used to limit the time step size, different processors
were sometimes using different time steps in their local computations. This screwed
up convergence checks int the nonlinear operator such that only some processor thought
convergence was already achieved while others wanted to do more iterations.
With this commit the method now returns whether there is any well on any processor
with the constraint being true.
This commit adds an option that allows to enable detecting parallel
logging fallout (option --enable-logging-fallout-warning). It is now
false by default (previous behavior was as if the option was true).
If option is true a warning will be printed for any process with nonzero
rank that does try to log to *.PRT or *.DBG and the logged output will be appended
to these files at the end.
Previously runs of flow started with 'flow base.' Would result in base..DBG
files; note the double dot and the case sensitivity. Now the file will be
BASE.DBG in this case. Note that any extension will be stipped (not only
.DATA as before).
Previously all processes reported
Warning: Fast restart using SAVE is not supported. Standard restart file is written instead.
Now this is done only on the master process where logging is activated.
The problem was only present if UMFPack was not found and therefore
FLOW_SUPPORT_AMG was defined to true. In that case we experienced compile
errors in a source branch that would never be executed. Therefore we remove
the code there and throw an exception.
Previously the regex issued warning for all files containing ".[0-9]*." in the name
in the output directory. That was too general and created more warnings than we wanted.
This fixes fallout from the merge of PR #1512. Since then the
ParallelFileMerger was always given the current directory even if
another output directory was requested. This resulted in
CASENAME.<procid>.<ext> files in that directory never being merged and removed.
Now the ParallelFileMerger gets the output directory path again.
It should be --milu-variant and is it now. Maybe the parameter system
could be a bit smart and detect consecutive upper case letters and treat
them correctly.
We introduced two runtime parameters for this: ilu_redblack and
ilu_reorder_spheres. If the last one is false, we try to preserve
the ordering within in the colors. Otherwise we try to achieve a D2
(alternative diagonal) ordering.
These versions are inspired by the ones used in SuperLU and the enums
to choose them have simuilar names, but without leading S (MILU_1-MILU_3).
The following variants are supported (chosen by the enum MILU_VARIANT):
ILU: plain ILU
MILU_1: lump diagonal with dropped row entries.
MILU_2: lump diagonal with the sum of the absolute values of the dropped row
entries.
MILU_3: if diagonal is positive add sum of dropped row entrires. Otherwise substract them.
MILU_4: if diagonal is positive add sum of dropped row entrires. Otherwise do nothing
These versions are inspired by the ones used in SuperLU and the enums
to choose them have simuilar names, but without leading S (MILU_1-MILU_3).
The following variants are supported (chosen by the enum MILU_VARIANT):
ILU: plain ILU
MILU_1: lump diagonal with dropped row entries.
MILU_2: lump diagonal with the sum of the absolute values of the dropped row
entries.
MILU_3: if diagonal is positive add sum of dropped row entrires. Otherwise substract them.
MILU_4: if diagonal is positive add sum of dropped row entrires. Otherwise do nothing
We did add the dropped matrix blocks to the diagonal blocks,
but this is not MILU where oone only modifies the diagonal.
With this patch we fix this behaviour and now only modify the
diagonal of the diagonal block.
Using the parameter ilu_milu=true|false (default=false) the user can now choose
to use the modified ILU0 decomposition. If selected values will
not be dropped for nonzero entries but added to the diagonal of U.
This approach will result in A*e = L*U*e for vector e with all entries
beging 1.
matrix_add_well_contributions=true will add to both the preconditioner matrix
and the matrix of the linear operator. preconditioner_add_well_contributions=true
will only add the contributions to the preconditioner matrix.
It turned out that applying the well part of the full matrix has to be
done after the application of the non-well interactions. Otherwise we
screw up the ordering so much that convergence suffers a lot.
Kudos got Atgeirr for inspiration based on his testing.
It queries the Well whether the jacobian also contains well contributions.
If not then it applies them in the operator in addition. Thus the
well knows whether that is needed or not.
Clearing the auxiliary modules will result in the sparsity pattern being
recomputed. Previously we did this before each nonlinear solve. But for the
master branch the sparsity pattern was only computed once for the whole
simulation.
To get the same behaviour (one sparsity pattern computation) we change the
auxiliary module to include all well perforations that might become active
during the simulation and do this once up front of a simulator run.
Please note that the new matrix entries might also improve convergence for
the ILU if the well is not active.
Previously, we did this during assemble. Unfortunately, this resulted
in the sparsity pattern being recomputed every time instead of just
once for a non-linear solve.
This is only done upon request and uses the auxiliary module approach
provided by ewoms.
In the case of adding the influences we do not execute applyWellModelScaleAdd
or applyWellModel in the operator
Currently we run into issues with the parallel AMG if
redistribution happened on the coarsest level. That is
not detected by AMG and it will construct smoothers on
all processors present before the redistribution. Some of
them will get OwnerOverlapCommunication objects that have
an invalid MPI communicator in them (MPI_COMM_NULL) and an
MPI_ERROR will be raised as we communicate in the constructor.
With this patch we detect this situaton and set the pointer to
OwnerOverlapCommunication to null to prevent communication. A
sanity check that the matrix has zero rows has been added.
Markus Blatt