Merge branch 'master' into pr358/buptphyphy/addmatcreatenormalhermitian/master

PCBDDC: usage of benign trick even without change of basis on velocities

this could happen if the user has already changed the basis on the velocities for local matrices before calling KSPSetUp

Merge branch 'tisaac/dt-fix-geom-punning' into tisaac/plex-feature-tree-fv

Pull in a fix to FE geometry that affects the complex,64bit case

* tisaac/dt-fix-geom-punning: (240 commits)
Plex: fix t

Merge branch 'tisaac/dt-fix-geom-punning' into tisaac/plex-feature-tree-fv

Pull in a fix to FE geometry that affects the complex,64bit case

* tisaac/dt-fix-geom-punning: (240 commits)
Plex: fix type punning problem stuffing (PetscFECellGeom *) in a Vec
PGI compiler cannot handle [restrict] usage therefor disable restrict for PGI compiler Fix ex48.c to use PETSc standard way of marking restrict
PETSC_UNUSED must occur before the typename. With PGI compilers they error if it is after the type name
on solaris in TSAdaptChoose_Basic() negative root for zero could produce -infinity thus causing time step to be decreased even though the error estimate was zero
MatGetSchurComplement: revert test from LUMP back to DIAG and update output
there are no tests for opencl or viennacl so prevent warning message: make[2]: *** No rule to make target `testexamples_OPENCL', needed by `tree'. Stop. make[2]: *** No rule to make target `testexamples_VIENNACL', needed by `tree'. Stop.
incorrect initial conditions function that could produce floating point exception value (and does on windows) due to squareroot of non-negative number
forgot to fix rule for building ex194
move ex193.c to ex194 because someone else created a 193
remove outdate manual page comment
add missing VecRestoreArray()
add missing VecRestoreArray()
missing restore array in example
formating for print statements of test example with __float128 precision
fix example for __float128 precision by using PETSc math macros and casting PetscReal in print statements
remove Tao example custom monitor that produced different results for __float128 precision
all PETSc built in monitors require a valid PetscViewer, NULL is not acceptable
PETSc built in monitors always require viewers to be passed in, no longer support a NULL viewer
use tigher haptol for minres for quad precision because otherwise false positive of indefinite matrix appeared in example
check for PetscMPIInt overflow in ISAllGather()
...

show more ...

Merge branch 'master' into jed/process-options-handler

Merge branch 'tisaac/dmplex-stable-tet-refinement' into tisaac/plex-feature-tree-fv

This feature branch also changed the topology of nonconforming examples:
I am merging so that I can make the regre

Merge branch 'tisaac/dmplex-stable-tet-refinement' into tisaac/plex-feature-tree-fv

This feature branch also changed the topology of nonconforming examples:
I am merging so that I can make the regression output that should be
correct in master and next

* tisaac/dmplex-stable-tet-refinement: (594 commits)
plex: added tet shape regression test
plexrefine: remove replaced code
plex: update test output for new tet refinement order
plex: fix tet refinement for stable element shape
DMPlex: added test of shape quality to ex1
Better to use the proper error handling version of PetscViewerASCIIGetStdout() instead of PETSC_VIEWER_STDOUT_()
Generate a very useful error message if PetscLogView(), PetscLogEventGetPerfInfo() or PetscLogEventGetFlops() is called with a preceeding call to PetscLogBegin() and friends (also -log_summary which triggers a PetscLogBegin())
fix compile warnings
Plex: Fix bug with bcComps initialization
Turn off elemental shared libraries if --with-shared-libraries=0 is used
Bib: Added refs
Bib: Fixed a bunch of errors
add warning/error flag for '-mat_superlu_dist_parsymbfact' when MatInputMode=GLOBAL
makefile: fix format
tests: add back ex6f, runex6f targets - that got removed in 2f1d3d565a6a43067e6df75b043e6343b05fcd1d
accessing option that may never have been set
typo in DMNetworkSetNumVariables() reported by Sean Shahkarami
superlu_dist: update to using latest 4.1 version
fix compile warnings
Generate an error message when trying to overwrite default options of CN and BEULER
...

show more ...

PCBDDC: Workaround for rhs in benign subspace

When the rhs is already in the benign space for Darcy, CG suffers a lot (still need to write down some formulas to understand why); instead GMRES conver

PCBDDC: Workaround for rhs in benign subspace

When the rhs is already in the benign space for Darcy, CG suffers a lot (still need to write down some formulas to understand why); instead GMRES converges in 1 (!) iteration.
With this, CG converges without iterating

show more ...

PCBDDC: nullspace info for the benign trick

PCBDDC: remove unneeded lines of code

PCBDDC: extend benign trick to PCApplyTranspose_BDDC

PCBDDC: make code for checking operator a function

PCBDDC: make code more clear

PCBDDC: move check into separate function

PCBDDC: move code to a separate function

PCBDDC: use consistent name for setting or retrieving P0 values

PCBDDC: Enable benign trick even with constraints

The user is responsible to either

1) use an iteration matrix whose basis is already changed to make \int_F v*n a primal dof for all F
or
2) attach

PCBDDC: Enable benign trick even with constraints

The user is responsible to either

1) use an iteration matrix whose basis is already changed to make \int_F v*n a primal dof for all F
or
2) attach a proper MatNullSpace object which guaranties that B0_R is null
or
3) provide a proper change of basis

In cases 1) and 3), adaptive selection of constraints could be used.

show more ...

PCBDDC: avoid playing with pointers. Use API instead

PCBDDC: remove primal vertices from face/edge deluxe scaling

this does not change the behaviour of standard BDDC.
It is required by the benign trick in order to
keep the iterates in the benign space

PCBDDC: remove primal vertices from face/edge deluxe scaling

this does not change the behaviour of standard BDDC.
It is required by the benign trick in order to
keep the iterates in the benign space

remove extra code to check benign-ness inside PCBDDCMatMult_Private

show more ...

PCBDDC: raise error if benign trick and user defined change of basis are both present

PCBDDC: Wrong update during postsolve and forgot to restore the original local matrix

Now the benign trick works
Still need to do something on the deluxe side

PCBDDC: default values should be set during creation

PCBDDC: added benign correction during presolve and check on the BDDC operator

still does not work, need to understand what is missing

there is some extra code that should be removed from PCBDDCMat

PCBDDC: added benign correction during presolve and check on the BDDC operator

still does not work, need to understand what is missing

there is some extra code that should be removed from PCBDDCMatMult_Private

show more ...

PCBDDC: remove zerodiag from bddc data structure

Merging with Barry's refactor of PetscViewerGetSubComm

Merge branch 'barry/fix-petscviewer-attempt-2' into dmay/pcsemiredundant

Merge branch 'barry/fix-nonew-notcollective/maint' into jed/mat-assembly-perf

Jed this fucking 6+ month running thing in next that is not in master is a royal fucking pain in the ass and
an abuse of

Merge branch 'barry/fix-nonew-notcollective/maint' into jed/mat-assembly-perf

Jed this fucking 6+ month running thing in next that is not in master is a royal fucking pain in the ass and
an abuse of git. I think it is not justified to ever have anything in next for more than a few weeks at most.
Either take it out of next if it is broken or put it in master if it is not broken. Hanging around in next but
not master for ever is not a good policy.

show more ...

PCBDDC: added mechanism to pass p0 from residual vector to coarse space

still does not work completely
need to add more checks