Actually communicate the shared memory node vector values through the shared memory region
This requires keeping a seperate set of indices from which to copy from and to for the shared memory region

Actually communicate the shared memory node vector values through the shared memory region
This requires keeping a seperate set of indices from which to copy from and to for the shared memory region
Note the code still sends and receives all data including that is duplicatively copied with the shared memory

Funded-by: P-ECP
Time: 1 hour

show more ...

Completed determination of shared memory partner VecScatter locations include addition of code to
move data across shared memory (currently still also sends data via messages).
Runs clean with valgri

Completed determination of shared memory partner VecScatter locations include addition of code to
move data across shared memory (currently still also sends data via messages).
Runs clean with valgrind

Funded-by: P-ECP
Time: 2 hours

show more ...

Successfully pack and then unpack the data in the shared memory region with VecScatter

Time: 5 hours
Funded-by: P-ECP

minor progress in begining to penetrate vec scatter code

Complete benchark example with display of logging communication time for VecScatter

Documented binding options for Intel mpirun

Funded-by: P-ECP
Time: 4

Merge branch 'master' into knepley/feature-plex-adjoint-prelim

* master: (60 commits)
petsc4py: update to use latest 3.8 snapshot
petsc4py: update to snapshot that syncs with 276c550
fix warni

Merge branch 'master' into knepley/feature-plex-adjoint-prelim

* master: (60 commits)
petsc4py: update to use latest 3.8 snapshot
petsc4py: update to snapshot that syncs with 276c550
fix warning
Plex ex26: Protect ExodusII stuff
DM: DMCreateSubDM() field argument should be const
TS ex46: Newlines and some numerical jitter
TS ex45: Newlines and some numerical jitter
TS ex11: Fixed indentation and time step numbering
SNES ex12: Update for DM view changes
Plex tests: Update to new DM output
Plex ex23: Use Nc not dim for vector fields
Plex: Default section now ignores cells hanging off boundary meshes
Plex: Viewer now reports the cell Height
Increase patchlevel to 3.8.3
Plex: Fixed DMPlexInterpolate() bugs - Do not copy associated data if we are not interpolating - Do not interpolate if we already have full depth
fix valgrind errors with using PetscObjectGetName()/SetName() with the same object - resulting in a free() before copy.
ex4: fix compile warning and update makefile
Minor fixes
Improving report_tests
PetscDS: Must initialize the default adjacency
...

Conflicts:
include/petsc/private/dmimpl.h
include/petscdm.h
include/petscis.h
src/dm/impls/plex/plexcreate.c

show more ...

fix random/ex2 (complex) and missing output file warning

DM: DMCreateSubDM() field argument should be const

Section: Fix init in PetscSectionCreateSupersection()

Section: If global section allocates bc indices, must copy them

Section: Cleaned up PetscSectionCreateGlobalSection()
- Used local variable and added input checks

convert examples to new harness that have depends: on other source code

Commit-type: testing-fix

Merge branch 'barry/test-harnes-vec-tests' of bitbucket.org:petsc/petsc into barry/test-harnes-vec-tests

convert a couple more examples over to new test harness

Commit-type: testing-fix

Convert another directory to new test harness

Commit-type: testing-fix

Converted over another test directory to new test harness

Commit-type: testing-fix

convert another directory to new test harness

Commit-type: testing-fix

converted directory to new test harness with automatic system

Commit-type: testing-fix

Merge branch 'master' into knepley/feature-plex-adjoint-prelim

* master: (251 commits)
yaml: add a backup download URL
Fix documentation of MatSetRandom function
Ignore -malloc-hbw option if m

Merge branch 'master' into knepley/feature-plex-adjoint-prelim

* master: (251 commits)
yaml: add a backup download URL
Fix documentation of MatSetRandom function
Ignore -malloc-hbw option if malloc is already set
Plex: Fixed array initialization
pgi: compile fix
knl: update test to use memkind and remove option that worked arround a bug with intel 17
Check the return value for memkind_posix_memalign
Change the default behavour for -malloc_hbw
Bib: More refs
test: remove runex13f90 as its now converted to new test suite
fix win32 crash. This test was previously run with '-n 4' - so restoring it
Fix petscsfgetgraph_ to work for all Fortran compilers by directly using array of PetscSFNodes
bugfix for MatDuplicate_MPIAIJ_PtAP Reported-by: Samuel Lanthaler <s.lanthaler@gmail.com>
strip out dead sundials usage code; change sundials test to only run 5 timesteps to prevent different number of steps on different systems.
Remove extra lines of output file
Example uses C99 constructs so must mark with requires: c99
fix const missmatch for DMDAVecRestoreArrayRead()
remove conflict between function name in example and new PETSc TS function
builds: attempt to do maint/master/next test builds in separate clones
ex62: it turned out that CHACO partioning is not deterministic. Use a prescribed partition instead
...

Conflicts:
src/dm/interface/dm.c
src/snes/examples/tutorials/ex13.c

show more ...

converted to new test harness with automatic system, only one issue reported directly to Scott

Commit-type: testing-fix

Merge remote-tracking branch 'origin/barry/vec-to-test-harness'

Merge remote-tracking branch 'origin/barry/fortran-PetscSFBcastBegin'

pgi: compile fix

Merge remote-tracking branch 'origin/pr803/acroucher/acroucher/fix-IS-global-to-local-mapping-block/master'

Fix petscsfgetgraph_ to work for all Fortran compilers by directly using array of PetscSFNodes

Commit-type: bug-fix, portability-fix, testing-fix
Reported-by: nightly tests