Users of PETSc,
We are looking for images of simulations (and shorted movies) that benefited from the use of PETSc for our website as well as our various programming managers who provide support for the development of PETSc. If you have some images we’d appreciate you sending them or a link to them to petsc-maint(a)mcs.anl.gov with a very short description, maybe a URL, and your name and institution information.
Thanks
Barry
You can also send us additions to our partial list of publications that use PETSc at http://www.mcs.anl.gov/petsc/publications/index.html
Dear PETSc users,
The patch release petsc-3.5.2 is now available for download.
http://www.mcs.anl.gov/petsc/download/index.html
Some of the changes include:
* ml: fix partitioning with parmetis
* build: fix gnumake dependencies and verify minimum required gnumake version.
* fix: ISLocalToGlobalCreateIS() for bs > 0
* fix: ISLocalToGlobalMappingCreateIS() to respect the block size of the IS
* KSPSetSupportedNorm: simplify priority logic around KSP_NORM_NONE
* AIJ/BAIJ: Performance improvement for MatZeroRows(), MatZeroRowsColumns()
* PetscFV: Fully enable other limiters
* TS+Plex: Overlap cells were misidentified as ghost cells
* fix: MatAXPY_ for mpibaij, mpisbaij, seqbaij and seqsbaij
* pastix: updated to version 5.2.2 This prevents crashes with diagonal matrices
* DMDA: Fix VecView() for two-dimensional grids with curvilinear coordinates
* DMDA: Fix VecView() displaying of coordinate limits in parallel
* configure: use LIBS option verbatim in the link line.
* configure: check if the compiler returns zero error code on link failure
Satish
Dear PETSc users,
The patch release petsc-3.5.1 is now available for download.
http://www.mcs.anl.gov/petsc/download/index.html
Some of the changes include:
* PCApplyRichardson_MG() did not properly set operators for all levels for residual computation Needed the same code as PCApply_MG() had
* if vector is of length 0 then VecSet() should set all norms to zero instead of triggering a divide by zero
* allow complex numbers to be used from both C and C++ (when supported) not only the language PETSc was built with
* improve the memory usage of PCGAMGCreateGraph()
* did not mirror 1d DMDA values correctly due to previous change consulidating block size -options_table should have been -options_view twenty years ago
* mpi4py: use current tarball mpi4py-1.3.1.tar.gz
* bugfix for memory leak with >1 processes and calling KSPSetOperators(A) twice with same A (see Jose Roman's email on Jul 11, 2014)
* At configure time determine if MPICH_NUMVERSION or OMPI_MAJOR/MINOR_VERSION or MPI-Uni, then at compile time make sure versions match to determine if incorrect mpi.h is being included
* petsc4py.py: look for python dynamic library in correct locations on Apple based on python version Test did not work correctly for anaconda installed python on Apple since that does not install a generic libpython.dylib
* Added code to properly preallocate "diagonal" portion of scalar matrix derived from vector matrix in PCGAMGCreateGraph()
* KSPCreate() for TAO GPCG solver is now done at TaoCreate_GPCG() instead of TaoSetUp() this allows users to change KSP options from the code easily and matches the style of other Tao solvers that use KSP
* TAO: Enable interface for complex scalar builds
Satish
We are pleased to announce the release of PETSc version 3.5 at http://www.mcs.anl.gov/petsc
The major changes and updates can be found at
http://www.mcs.anl.gov/petsc/documentation/changes/35.html
We recommend upgrading to PETSc 3.5 immediately. As always please report problems to
petsc-maint at mcs.anl.gov and ask questions at petsc-users at mcs.anl.gov
Notable new features/changes include
* KSPSetOperators(), SNESSetJacobian(), TSSetRHSJacobian(), and TSSetIJacobian()
no longer have a MatStructure argument, PETSc automatically tracks if the matrix structure has changed.
The Mat * arguments in these functions have also been changed to Mat
* the Tao scalable optimization software package is now included directly in the PETSc releases
* added DMNETWORK for simulations/computation on graphs, networks, and circuits
* DMDA requires much less memory usage in most situations
* --download-suitesparse now supports all solvers in suitesparse and replaces —download-umfpack and —download-cholmod
* added support for MKL’s Pardiso with ./configure --with-mkl_pardiso-dir and -pc_factor_mat_solver_package mkl_pardiso
* compiles of PETSc source now, by default, use GNU Make instead of Cmake and are more robust
* many addition/improvements to DMPlex unstructured mesh management
* there is no longer separate ISLocalToGlobalMappingBlock objects, instead the ISLocalToGlobalMapping
contains a block size
If you are using the threaded version of PETSc or PETSc on GPUs you should continue to work with
petsc-dev since this code is too much in flux to be contained in a release.
This release includes contributions from
Andy R. Terrel
Barry Smith
Blaise Bourdin
David Ham
Debojyoti Ghosh
Dmitry Karpeev
Elliott Sales de Andrade
Emil Constantinescu
Evan Gawlik
Geoffrey Irving
Geoffrey Oxberry
Hong Zhang
Jack Poulson
Jason Sarich
Jed Brown
Jose David Bermeol
Jose Roman
Juha Jäykkä
Jungho Lee
Karl Rupp
Lawrence Mitchell
Lisandro Dalcin
Mark Adams
Matthew Knepley
Michael Lange
Patrick Farrell
Patrick Lacasse
Patrick Sanan
Paul Mullowney
Peter Brune
Richard Tran Mills
Rémi Lacroix
Satish Balay
Shao-Ching Huang
Shri Abhyankar
Stefan Lemvig Glimberg
Stefan M. Wild
Stefano Zampini
Tobin Isaac
Todd Munson
Victor Eijkhout
Vijay Mahadevan
Wenjun Deng
Wolfgang Bangerth
and bug reports/proposed improvements received from
Alan Richardson
Alexander Krimm
Boris Kaus
Brad Aagaard
Christiaan Klaij
Dave May
David Ketcheson
Derek Gaston
Dharmendar Reddy
Einar Sørheim
Eloy Romero Alcalde
Garth N. Wells
Gautam Bisht
Ilmari Karonen
Jay Jay Billings
Kai Germaschewski
Lawrence Mitchell
Loris Bennett
Luc Berger-Vergiat
Lukasz Kaczmarczyk
LV CHAO
Mani Chandra
Matteo Parsani
Max la Cour Christensen
Michele De Stefano
Mikael Öhman
Negin Niknaam
Paolo Orsini
Patrick Farrell
Patrick Lacasse
Paul Bauman
Pratheek Shanthraj
Randall Mackie
Richard Katz
Tabrez Ali
Torquil Macdonald Sørensen
Vaclav Hapla
Vincent Frans Maria De Groof
Vitali Morozov
Xiangdong
As always, thanks for your support,
Barry
The Argonne Training Program on Extreme-Scale Computing for advanced graduate students, postdocs, and early career computational scientists and engineers who might profit from the program. This will have only a short presentation on PETSc and packages PETSc uses but it covers many issues related to parallel high performance computing.
See http://extremecomputingtraining.anl.gov registration closes soon.
Barry
Dear PETSc users,
The patch release petsc-3.4.4 is now available for download.
http://www.mcs.anl.gov/petsc/download/index.html
Some of the changes include:
* MatDense: fix overflow with int32*int32 during malloc()
* VecView_MPI_DA: fix gsizes bug (bad conversion in parent commit)
* MATSNESMF: fix so that MatMFFDSetFunction() works correctly with it
* configure: enable hdf5 even when its not built with a compression library
* configure: check for ddot_() in chaco and reject it - if found
* configure: print download package URL before downloading the package
* MatNullSpaceCreateRigidBody: fix array length for 3D
* MatMPIAIJGetLocalMat: Correctly check input type of Mat argument
* Matlab: Fix Matlab viewer for parallel complex vectors
* Mat: Turned on Fortran binding for MatSetValuesBlockedStencil()
* Mat: bugfix in MatCreateSeqAIJFromTriple()
* SNESLINESEARCHBT: Set the norms when exiting early due to negligible step
* configure: detect if winzip is used to extract petsc.tar.gz and error out
* configure: check for windows-python and give error message
* Mat: Check column block size in MatSetBlockSizes
* Mat: add non-square block support for MatSetValuesBlocked
* PetscSynchronizedFGets: fix deadlock at EOF
* Mat: fix bad LogFlops in MatSOR_SeqSBAIJ.
Satish
PETSc users,
We are please to announce the availability of SAWs, the Scientific Application Webserver, which easily makes any long running C program into a webserver allowing you to monitor the progress of your simulation with any browser. SAWs is open source (BSD license), small, portable and requires only a C compiler and Unix/Linux. SAWs and its documentation can be found at http://bitbucket.com/saws/saws. An example of a running application can be accessed at http://petsc.no-ip.org:8080
SAWs can be used standalone, independent of PETSc, or in the development version of PETSc we have code that works together with SAWs as demonstrated in the above mentioned web page.
Jed, Matt Otten, and Barry
Dear PETSc users,
The patch release petsc-3.4.3 is now available for download.
http://www.mcs.anl.gov/petsc/download/index.html
Some of the changes include:
* Sys: Fix C/C++ choice for PETSC_RESTRICT and PETSC_STATIC_INLINE.
* PCGAMG: fix corruption for multiple solves when MatSetNearNullSpace is used
* Mat: add MATOP_ values for missing functions
* Intel: Fix error encountered when using Intel compiler and changing gcc on
PATH between configuring PETSc and building an application.
* Configure: avoid string argument to __attribute((deprecated))
* KSPDGMRES: fix conditional variable declarations
* MatSBAIJ: avoid invalid memory read in MatSOR_SBAIJ
* BGQ: Work around BG/Q MPI_Bcast bug.
* PetscOptionsInsertFile: combine Bcasts for alias and values
* Sys: ex13 avoid non-portable realpath(1) in verification test
* Sys: fix PetscGetFullPath expansion of ~ and ~user, ensure NULL termination
* Sys: avoid getpwuid in file path routines
* MatXAIJSetPreallocation: use array[] notation to help out Fortran
* Cygwin64: Fix build error with fenv.h
* MatSchurComplement: Must initialize KSP package
* SNES: Fix accidental creation of the matrices in SNESNCG used for linear analogue LS.
* PCGAMG: add man page for PCGAMGType
* SNES: Fixed SNESComputeJacobianDefaultColor() in the case of multilevel PC
* Mat: fix Fortran interface for MatGetColumnNorms
* Bfort: distinguishes between array[] and *pointer to decide whether the
Fortran interface should have an array or a scalar.
* DM: Fix for accidentally duplicated refcount decrement in dm->x
* Doc: PetscLayoutCreate format code block and list
* Sys: add PetscDataTypeFromString() and test code
* PCGAMG: Always rebuild coarse Mat the first time
* PCFactor: add PCFactorGetLevels
* Dgmres: added option to view deflation vecs after solve.
* MatPCFactor: include MatFactorShiftType in PCView_Factor output
* AO: fix erroneous processing of -ao_view and factor into AOViewFromOptions
* Viewer ASCII: print tabs to petsc_history, not fd
* PCBJacobi: call KSPSetFromOptions(subksp) when PCSetUp was called first
* PCGAMG: set bjacobi->same_local_solves=TRUE to clean coarse grid viewing
* SNESMonitorCancel: enable fortran interface
* BGQ: fix deadlock on bgq due to inconsistant values returned by getpwuid(getuid())
* Scalapack: fix build by specifying mpi.h to build
* MatSeqAIJGetArrayF90: add man pages and fix typo in MatDense version
* SF docs: add link to summary page and fix title
* Doc: Fixed DM code in manual
* VTK: Small fix to error message (.vts to .vtu)
* Config cmakeboot: add C++ flags any time compiler is available
* Config OpenMP: check for C++ flag any time the compiler is available
* Parms: update tarball with fix for namespace conflict with metis
* Parmetis: update tarball to parmetis-4.0.2-p5 which fixes an install issue with cygwin
* Viewer: fix Binary Matlab viewer info to use '#' instead of '%'
* MatSetOption not always logically collective.
* SNESVI: Fix computation of squared-norm in for the bt linesearch in the VIRS snes case
* KSPSetDM docs: fix KSPSetDMActive args, add cross-references
* Vec: fix flop counting to use map->n instead of non-existent map->m
* Hdf5 installed without zlib should configure without neededing zlib dependency
* Config: name of configure generated include file confdefs.h did not match name printed in configure.log of conftest.h
* Vec: put back in flop counts for VecNorm_Seq
* KSPFGMRES: make KSPMonitor usage match KSPGMRES
* TS: Tutorial example has no embedded method so needs -ts_adapt_type none
added MatSetUp calls to TS tutorials 16 and 21
* DM: Updated DM test ex7.c to use valid command-line options to specify 2D and 3D grids.
* PCMG: Fixed Fortran binding
* TS: fix MSVC compiler error: non-POD member of anonymous struct
* TS tests/ex3: SNESFD and TSComputeIJacobianConstant are incompatible
* TS: fix shortcut scaling; document TSComputeIJacobianConstant limitations
* TS: add TSRHSJacobianSetReuse, fixes TSComputeRHSJacobianConstant
* Zoltan: X is listed as a required dependency of Zoltan but it is only optional
* Doc: add petsctime.h to SOURCEH (fixes missing docs)
* ABI: fix ABI compatibility for debug and optimized builds
* Clang: Fix warning with clang-3.3.
* Mat Chaco: remove another case of PETSC_DEFAULT out of range of enum
Satish
Dear PETSc users,
The patch release petsc-3.4.2 is now available for download.
http://www.mcs.anl.gov/petsc/download/index.html
Some of the changes include:
* plog: Fixed Fortran binding for PetscLogPrintDetailed()
* fixed src/dm/examples/tutorials/ex7.c to use replacement for PetscViewerBinaryMatlabOpen() also needed to fix bin/matlab/PetscBagRead.m for new location of bagimpl.h and put bagimpl.h into the appropriate public place
* badly formatted /*I "include" I*/, unused variables fixed that compile with MATLAB
* sundials: disable f77 properly
* exodus: cleanup build process [make.inc is unused & fix postInstall() to use the correct file]
* exodus: do 'make clean' between repeat builds
* mpicc: disable 'wrapper override check' when user provides --with-mpi-compilers=0
* Fortran: PCMGResidual_Default needs to handle double-underscore
* version: fix PETSC_VERSION_LT() to require PETSC_VERSION_RELEASE=1
* Escape quotes in paths in petscmachineinfo.h
* configure: handle paths with spaces in getExecutable(). Now the following works: '--with-mpiexec=/cygdrive/c/Program\ Files/MPICH2/bin/mpiexec -localonly',
* tgamma: icc on windows requires mathimf.h for tgamma. [but configure doesn't use any includes in tgamma test]
Satish
Dear PETSc users,
The patch release petsc-3.4.1 is now available for download.
http://www.mcs.anl.gov/petsc/download/index.html
Some of the changes include:
* KSPPIPECG: bug fix for PRECONDITIONED and UNPRECONDITIONED norms
* Clean up communicator attribute deletion (works around bad MS-MPI attribute handling)
* Remove reference to PETSC_COMM_WORLD from DMPlex and TSARKIMEX
* SNES: bug fix when MatFDColoring is configured by user
* SNESFAS: allow SNESView before setup
* HDF5: add missing Fortran bindings
* Xcode support restored
* A few documentation fixes and improved diagnostic output
Satish
