merge release-3.1 (MPI_Win_fence guard)

Hg-commit: 978e67a54309b1ef45a156b36999684ffaaa38d1

Guard MPI_Win_fence with PETSC_HAVE_MPI_WIN_CREATE

Hg-commit: 78047f5fe15612d08bb0543c091e90ec256f96a7

merge release (-vecscatter_window)

Hg-commit: 33119f34691172a04eb8b481681635d0a70ac063

Bug fixes for -vecscatter_window

* The MPI_Put was not using winstarts, meaning that every process
dropped their result in the same slot (incorrect results with np>2).

* Block size was not used f

Bug fixes for -vecscatter_window

* The MPI_Put was not using winstarts, meaning that every process
dropped their result in the same slot (incorrect results with np>2).

* Block size was not used for starts, so the wrong data was getting sent
and the window was too small for the packing when bs>1.

* Enhancement: the closing MPI_Win_fence has been moved to VecScatterEnd
so that there is some room for the Puts to be processed while
computation occurs.

* Added -vecscatter_window to snes runex19_5 (where all the other
vecscatter options are tested).

Hg-commit: 7607898b3ff596e346f2642260abd49b23b5aa25

show more ...

commit after merge

Hg-commit: 78a971ae3462970fe26653af75a34c7f98a831bc

added XXXSetDM() and XXXGetDM() for PC,KSP,SNES and TS

Hg-commit: bada9ff4302cce1dee88b5dd282aa59af62f2a84

use extern_c for fortran kernels

Hg-commit: 30f20350e1dcbc1a8f90d63ba553086609a4dd87

more windows compile fixes

Hg-commit: 7ed438e12a20c0b2931ca14dbfc128900a59415b

fixed false possible use of uninitialized variables warnings with optimization

Hg-commit: 4ff7b4bc0bea1f05f5ba7c1668b55e0b648cb6a9

clarify wording in ISLocalToGlobalMappingGetInfo

Hg-commit: 483941e350f3e39e4074d00711367d76ee0cc11e

Fixes for -vecscatter_window

Hg-commit: 6ae96dca2adc4ce63991647a74bd9dbf9ccb923d

moved use petsc and use petscdef to use petscsys and use petscsysdef

Hg-commit: b73864045211bef46a51b4cf8f1679a8858a6e42

fixed some dead code objects that prevents Matt's builder.py from completing.
added hacks to Matt's builder.py to get success on my laptop.

Hg-commit: 893fc1a80ebe376870cac09c70bb9806e6d51cae

only files that need it get petscblaslapack.h included because it conflicts with prototypes included with some external packages.

Hg-commit: 9b075c74df17031ee2599abf4807903f0312acd0

petscblaslapack.h now included by petscsys.h
some fixes to builder.py

Hg-commit: 56bbc3637d96c6c99b427f2d14f38bcf58262255

fix for PaStix.py download for Mac

Hg-commit: b1ed3fb65d1d6ad29529b7346c7183d9e8adefcf

petsc.h --> petscsys.h (fortran equivalents also)
new petsc.h that includes all PETSc includes is added

Hg-commit: a82a9bcb80ed39c69991db706672838a5b792d44

moved petscsys.h stuff into petsc.h files (Fortran and C) removed petscsys.h files fortran and C
eliminate use of petscsys.h (Fortran and C)
eliminated redundant use of #include "petscfix" all over t

moved petscsys.h stuff into petsc.h files (Fortran and C) removed petscsys.h files fortran and C
eliminate use of petscsys.h (Fortran and C)
eliminated redundant use of #include "petscfix" all over the place because included by petsc.h

Hg-commit: dea4bc8a57aa0490c3a40fdee08c4b41e42e03ce

show more ...

fixed lsqr code to compile
added flowcontrol variable to PetscViewer Binary and took out need for HANDSHAKE macro

Hg-commit: ff43b0762f3f6bda026f3bd49b1b6267a30547e0

changed PetscSum_Op to MPIU_SUM to match the style we have for extra MPI stuff

Hg-commit: 6f826cf21d32c05c8a481b7d8be71073e672cf40

ISAllGather: free Malloc2'd memory with Free2

Hg-commit: a181b5fcff6a724fed887f254e76d7eaa88b9a9d

Would someone please kick the jerk who neglected to put __func__ into C89

Hg-commit: a6689f03dc2800f8f5cbf8be70469a56f89e8bee

Add ISCopy() to IS interface.

Has the same semantics as VecCopy, etc.

Hg-commit: 0b49cefef6c36f13059768efe00e72b7aeadf665

Quoting in VecLoadIntoVector man page

Hg-commit: bcb83e8f9398f5cbb316d59011c5227975283e32

correcting of 0 assignments to double to be 0.

Hg-commit: 4e8d3c66ce68ac50361b8dc9292a6438842e1543