mpich: revert to mpich-3.2 on osx and linux on PGI

On some OSX builds (osx-xsdk-opt, osx-pgi-pkgs-opt but not all?) we get MPIIO errors with src/vec/vec/examples/tests runex46_mpiio etc
******* Tes

mpich: revert to mpich-3.2 on osx and linux on PGI

On some OSX builds (osx-xsdk-opt, osx-pgi-pkgs-opt but not all?) we get MPIIO errors with src/vec/vec/examples/tests runex46_mpiio etc
******* Testing: testexamples_C *******
1,11c1
< *** PetscViewer[write] using MPI-IO ***
< *** PetscViewer[write] skipping header ***
< *** PetscViewer[load] skipping header ***
< *** PetscViewer[load] using MPI-IO ***
< VecCompare
< min(a) = +8.40e-04 [loc 8]
< max(a) = +9.95e+03 [loc 3]
< min(b) = +8.40e-04 [loc 8]
< max(b) = +9.95e+03 [loc 3]
< min(a-b) < 1.0e-10
< Headerless read of data verified
---
> Warning: Executing TestBinaryMPIIO() requires a working MPI-2 implementation
/Users/petsc/petsc.clone-4/src/vec/vec/examples/tests
Possible problem with with ex46_mpiio, diffs above
=========================================
26a27,29
> WARNING! There are options you set that were not used!
> WARNING! could be spelling mistake, etc!
> Option left: name:-viewer_binary_mpiio (no value)
/Users/petsc/petsc.clone-4/src/vec/vec/examples/tutorials
Possible problem with ex5_1, diffs above
=========================================

On linux with PGI 16.4 - there is build error with:
/usr/lib/gcc/x86_64-linux-gnu/4.8/libbacktrace.a(backtrace.o): In function `unwind':
(.text+0x16): undefined reference to `_Unwind_GetIPInfo'
/usr/lib/gcc/x86_64-linux-gnu/4.8/libbacktrace.a(backtrace.o): In function `backtrace_full':
(.text+0xa1): undefined reference to `_Unwind_Backtrace'

show more ...

mpich: update to version 3.3a2

Merge branch 'master' into knepley/feature-parallel-mesh-load

* master: (180 commits)
Bib: Added ref
-random_view caused error because PetscRandomView() was not provided in PetscRandomCreate()

Merge branch 'master' into knepley/feature-parallel-mesh-load

* master: (180 commits)
Bib: Added ref
-random_view caused error because PetscRandomView() was not provided in PetscRandomCreate()
Update the xSDKTrilinos test message to use interactive session and ctest
gcc6: fix comple warnings with -Wmisleading-indentation
fix compile warning
bugfix for MatTransposeMatMultSymbolic_MPIAIJ_MPIAIJ_nonscalable()
minor formatting cleanup of gamg.c
add ml example to tutorials page for xsdk
bugfix of MatDuplicate for mpiaij matrix product Reported-by: Miorelli, Federico <Federico.Miorelli@cgg.com>
The deprecated -log_summary_exclude was still being used when it should have been -log_exclude
cpardiso: requires mkl_blacs so check for that. Also common code to detect MKL for both pardiso and cpardiso
cpardiso: fix segv and memory leaks
resume MatGetSubMatrix_SeqSBAIJ() because it cannot share routine with BAIJ due to cast
fix bad relative links
replace MatSeqBAIJSetPreallocation_SeqBAIJ with MatSeqBAIJSetPreallocation to skip this call when input mat is sbaij type
fix a variety of set but not used variables in examples detected by clang static analyzer
removed declared but not used variable
ANL Report formatting requires listing authors by affiliation
DMP4est: uncrustify
DMP4est: remove memset()
...

show more ...

Merge branch 'master' into stefano_zampini/feature-pcbddc-saddlepoint

Conflicts:
src/ksp/pc/impls/bddc/bddc.c
src/ksp/pc/impls/bddc/bddcgraph.c
src/ksp/pc/impls/bddc/bddcprivate.c

Attempt to fix --with-batch to work reasonably with --download-xxx

Merge branch 'master' into stefano_zampini/feature-pcbddc-saddlepoint

Conflicts:
src/ksp/pc/impls/bddc/bddc.c
src/ksp/pc/impls/bddc/bddcgraph.c
src/ksp/pc/impls/bddc/bddcprivate.c
src/ksp/pc/imp

Merge branch 'master' into stefano_zampini/feature-pcbddc-saddlepoint

Conflicts:
src/ksp/pc/impls/bddc/bddc.c
src/ksp/pc/impls/bddc/bddcgraph.c
src/ksp/pc/impls/bddc/bddcprivate.c
src/ksp/pc/impls/bddc/bddcscalingbasic.c
src/ksp/pc/impls/bddc/bddcschurs.c
src/mat/impls/is/matis.c

show more ...

Merge branch 'master' into barry/add-pcfieldsplit-logevent

Merge branch 'master' into tisaac/dmforest

* master: (87 commits)
fixed --download-xxx=/directory to not mistake other options such as --download-mpich-pm=hydra etc
Correct spelling for Autotool

Merge branch 'master' into tisaac/dmforest

* master: (87 commits)
fixed --download-xxx=/directory to not mistake other options such as --download-mpich-pm=hydra etc
Correct spelling for Autotools libtoolize in commit a0848688a08a6eb150afdc95a2f6ce49cd4cd079
Fix for commit c8b462efbd8e8a5f06df500cf417f3a48f2b88d9 Put back the gitpreReq() check that was needed for MOAB
website: fix Fluidity link
bib/cite: add UMFPACK
nightlybuilds: update test tu use newer intel compiler [detect and use compiler path]
nightlybuilds: fix vec/ex7,ex7f link issue by using the correct LIBS option
Added support for users to provide their own download package installers --download-xxx=/directory where the directory contains the file xxx.py that is a standard PETSc package definition like in config/BuildSystem/config/packages/xxx.py
nightlybuilds: switch default intel compiler [& MKL] from 13.1.3 to 16.0.0.
mkl_pardiso: fix runex125_mkl_pardiso
mkl_pardiso: fix runex2_mkl_pardiso_cholesky
nightlybuilds: switch OSX to use boost from brew [as --download-boost is broken]
trilinos: gitcommit - always specify branch as origin/branch
Add error check to boost.py when running on Mac OS because the shared libraries created do not work since they do not have the proper rpath set
Update output files for mpi matrices that use to be seq
Move PostProcess() runs until after PETSc configure has saved all data to the filesystem so that make install will work correctly.
Switch to Trilinos release candidate for --download-trilinos Tested, failed with unresolved Kokos methods, Issue report to Trilinos team
Some configure packages UnitTest-CPP requires libtoolize to be called before autoreconf on some systems
--download-ideas now downloads unittest-cpp
simplied ./configure packages that need autoreconf to be run first (like moab and unittest-cpp) added unittest-ccp download changed downloadfilename to downloaddirname since that is what it is
...

# Please enter a commit message to explain why this merge is necessary,
# especially if it merges an updated upstream into a topic branch.
#
# Lines starting with '#' will be ignored, and an empty message aborts
# the commit.

show more ...

Merge branch 'barry/downloads'

simplied ./configure packages that need autoreconf to be run first (like moab and unittest-cpp)
added unittest-ccp download
changed downloadfilename to downloaddirname since that is what it is

Funde

simplied ./configure packages that need autoreconf to be run first (like moab and unittest-cpp)
added unittest-ccp download
changed downloadfilename to downloaddirname since that is what it is

Funded-by: IDEAS
Project: xSDK
Time: 3 hours

show more ...

Merge branch 'barry/downloads'

Merge branch 'master' into barry/add-pcreset-hypre-for-master

Needed to move PCDestroy_HYPRE() added to master to PCReset_HYPRE() since that was introduced
as a bug fix for maint

Merge branch 'barry/download-mstk' into barry/downloads

MSTK uses -DINSTALL_DIR instead of _DCMAKE_INSTALL_PREFIX, time spent debugging the problem
added simple rules for removing specific ./configure or cmake arguments from a list of arguments

Funded-by

MSTK uses -DINSTALL_DIR instead of _DCMAKE_INSTALL_PREFIX, time spent debugging the problem
added simple rules for removing specific ./configure or cmake arguments from a list of arguments

Funded-by: IDEAS
Project: xSDK
Time: 1 hour

show more ...

Merge branch 'master' into tisaac/dmforest

* master: (143 commits)
Revert "revolve: c++ compile fix" - as revolve now does not require 'bool' anymore
revolve: c++ compile fix
revolve: enable t

Merge branch 'master' into tisaac/dmforest

* master: (143 commits)
Revert "revolve: c++ compile fix" - as revolve now does not require 'bool' anymore
revolve: c++ compile fix
revolve: enable tests and add to nightlybuilds
fix download url for revolve
remove outdated unused script
removed outdated use of the term petsc-dev for the development version of PETSc
streams: precompute NPMAX - and use this value. And now user can use it optionally as:
mpich: default to using ver 3.2 on OSX/ElCapitan as 3.1.3 does not build on Barry's laptop
Revert "mpich: update version to mpich-3.2"
superlu/dist: disable build on freebsd - as this test is with bsdmake [and superlu requires gnumake]
superlu/dist: enable in complex nightly builds
fortran: fix typo with MATSEQAIJRESTOREARRAYF90
Documentation for Dense and SeqAIJ Get/RestoreArrayf90 had incorrect dimensions
Docs for MatGetRowIJF90() cut and paste and never fixed
Bib: NPC refs
superlu: no longer requires xerbla slamch dlamch ? $ nm -Ao libsuperlu_dist_4.3.a |egrep \(xerbla\|slamch\|dlamch\)
superlu_dist: update to version 4.3
Bib: More of our refs
added whitespaces to align numbers in verbatim environment
Added verbatim environment to special tokens
...

show more ...

mpich: default to using ver 3.2 on OSX/ElCapitan
as 3.1.3 does not build on Barry's laptop

Revert "mpich: update version to mpich-3.2"

This reverts commit 94e11dd9c51ae5b3fd4117d2433750fe95461db9.

mpich-3.2 is not valgrind clean

Merge branch 'master' into tisaac/dmforest

* master: (177 commits)
DT ex3: The test should check the number of digits requested
DT ex3: One test fails in single precision
Plex: Do not automati

Merge branch 'master' into tisaac/dmforest

* master: (177 commits)
DT ex3: The test should check the number of digits requested
DT ex3: One test fails in single precision
Plex: Do not automatically create a default section if no fields are defined
Plex: DMPlexCopyCoordinates() needs to create the coordinate section if it is missing
superlu: no longer requires xerbla slamch dlamch ?
superlu: update to v5.1
skip MatSolve() if matrix factorization fails
replacing SETERRQ with error output info and continue execution
SNES ex12: Added parallel tests - These test parallel, unstructured interpolation/restriction using FAS
Plex+SNES: Now in DMPlexComputeJacobian_Internal(), we also form the Jacobian preconditioner matrix
FE: Now we can form either the Jacobian or the Jacobian preconditioned matrix using a flag
DS: Now have point wise function slots for the Jacobian preconditioned matrix - This will allow us to form the "right" operator for preconditioning, like M in the (2,2) block for Stokes
Plex: When swapping Plexes, also swap point SF
Plex: Fix parallel error in Injector
Mat+Plex: Added MatPreallocator which can be called with MatSetValues() and preallocates another Mat - This could have been another "mode" of Mat, but I thought this was cleaner - Use MatPreallocator to preallocator Plex interpolators
DT ex3: Change tolerance and eliminate tests which need MPFR from regular runs
superlu: fix complex build
superlu: update to using 5.0 tarball with the fix for -Werror=format-security
superlu: update to use tarball with -Werror=format-security fix
DM: Clone the coordinate DM instead of reusing in Clone()
...

Conflicts:
config/builder.py
src/dm/impls/plex/plexcreate.c

show more ...

Merge branch 'master' into hongzh/checkpointing

Manual merges for a couple of PetscOptions changes in master

Merge branch 'balay/update-mpich-3.2'

fix: 620d467- split checkDownload and move only download_cygwin part to configure()
Fixes build error on windows with MS compilers.

Merge branch 'balay/update-mpich-3.2'

Merge branch 'jed/workaround-mpich-naughty-ibarrier'

update to latest MPICH since early one won't compile on El Capitan

fixes to compile with complex
changed malloc to PetscMalloc() and provided the appropriate PetscFree()
moved variable declaration from middle of function to top of function