Merge branch 'maint'

Merge branch 'balay-to-maint-fix-vs2015-vsnprintf' into maint

Merge branch 'jed/mat-assembly-perf' of bitbucket:petsc/petsc

VecAssembly and MatAssembly now use a scalable exchange pattern based on
PetscCommBuildTwoSided. This feature can be controlled with th

Merge branch 'jed/mat-assembly-perf' of bitbucket:petsc/petsc

VecAssembly and MatAssembly now use a scalable exchange pattern based on
PetscCommBuildTwoSided. This feature can be controlled with the options

-vec_assembly_bts 0 or 1 (default 0)
-matstash_bts 0 or 1 (default 0)

The rationale is that the new implementation with scalable data
structures can be slightly slower than the old version at small process
counts. The default here could be changed to depend on the process
count (leading to possibly-confusing scaling performance diagnostics) or
the implementation could learn to take a fast path.

* 'jed/mat-assembly-perf' of bitbucket:petsc/petsc: (49 commits)
MatStash: fix -Wsign-compare by using size_t for loop index when max is also size_t
Sys BuildTwoSided test: fix for non-POD std::complex
MatStash: cast to satisfy non-structural MPI type tag check
mpiuni: fix compile error /sandbox/petsc/petsc.clone-2/arch-linux-uni/lib/libpetsc.so: undefined reference to `MPI_Type_create_resized'
Vec: Silence compiler warning
Vec: fix typo in comment
Sys: fix C89 compiler warning
VecStash BTS: fix block stash InsertMode accounting
VecStash BTS: fix indexing bug counting sends to rank 0
MatStash BTS: fix memory leak on MAT_SUBSET_OFF_PROC_ENTRIES
VecAssemblyEnd_MPI_BTS: fix donotstash code path
VecAssemblyEnd_MPI_BTS: fix C++ conversion to InsertMode
Sys: fix datatypes test using MPI_Type_create_resized
MatStash BTS: work around lack of offsetof() for non-POD (std::complex)
MatAssembly: move check for InsertMode consistency into MatStashScatter impls
MatStash BTS: fix memory leak on reassembly with MAT_SUBSET_OFF_PROC_ENTRIES
MatStash BTS: small simplification to row ownership calculation
MatStash BTS: add MAT_SUBSET_OFF_PROC_ENTRIES, impl with neighbor-only comm
MatStash: initial BTS (BuildTwoSided) implementation
MatStash: add extension point for new BTS implementation
...

show more ...

Merge branch 'maint'

hypre: update with "Bug fix for Parasails; src/ksp/ksp/examples/tutorials/ex10.c runex10_16"

label additional packages that have test cases in PETSc

BuildSystem setCompilers: use fPIC test flags when filtering shared flags

Some systems (FreeBSD) don't like it when use use a "-shared" flag with
non fPIC objects, even if the user does explicitly u

BuildSystem setCompilers: use fPIC test flags when filtering shared flags

Some systems (FreeBSD) don't like it when use use a "-shared" flag with
non fPIC objects, even if the user does explicitly use the plt table, so
we have to move that part of the test inside the fPIC loop as well.

show more ...

move logic for which packages have tests for tests that require DATAFILESPATH into each package file
instead of globally in regression.py

gmake4: disable '--output-sync=recurse' on OSX
http://lists.gnu.org/archive/html/help-make/2015-08/msg00012.html

BuildSystem: check for positiion-independent code after shared linker

The only sure way to know if -fPIC is working is to compile a shared
object with undefined symbols, but to find how to build a s

BuildSystem: check for positiion-independent code after shared linker

The only sure way to know if -fPIC is working is to compile a shared
object with undefined symbols, but to find how to build a shared
library, we need to be able to compile PIC objects. So we have to test
both at the same time (at least for one language: PETSc default C
language). Then, once we have found a working shared linker, we can
test the other languages to see if they can compile shared objects with
undefined symbols.

show more ...

BuildSystem: fix incorrectly popped language

configure: fix _vsnprintf() check on win/vs2015 - it requires <stdio.h> in test

configure: fix --prefix --download-f2cblaslapack bug

Reported-by: Gilberto Perez <gilberto.limon.perez@gmail.com>

Merge branch 'knepley/fix-configure-compiler-log'

error out if elemental being installed and cmake was not found (copied from metis.py)

PGI compiler cannot handle [restrict] usage therefor disable restrict for PGI compiler
Fix ex48.c to use PETSc standard way of marking restrict

push for C language for concurrencykit test

add specific link test for concurrencykit

remove default configure options that offend concurrencykit

Configure: Forgot to restore log before some early returns/exceptions

Configure: Was missing some log prints

cleanup warning message to user about optimal optimization flags
and call setup() on compilerOptions module, per Matt, so that warning message is printed to screen

display warning if one configures with --with-debugging=0 but does not define
the optimization flags
Currently the warning is not printed to the screen but only to the configure.log
file because of t

display warning if one configures with --with-debugging=0 but does not define
the optimization flags
Currently the warning is not printed to the screen but only to the configure.log
file because of the funky way compilerOptions.py is loaded by compilerFlags.py

show more ...

use -g option with optimized builds for gnu, clang, and intel
use -O3 option for clang optimized builds

fix MPICH.py to build debug version of MPICH if PETSc is in debug mode
by using MPICH ./configure option --enable-fast=no
remove code from compilerOptions.py that tried to handle this by forcing
PET

fix MPICH.py to build debug version of MPICH if PETSc is in debug mode
by using MPICH ./configure option --enable-fast=no
remove code from compilerOptions.py that tried to handle this by forcing
PETSc flags to have a -O0

show more ...