]>
https://gitweb.dealii.org/ - dealii.git/log
Timo Heister [Fri, 27 Jan 2017 18:19:35 +0000 (13:19 -0500)]
fix tests/fe/bdm_16
Denis Davydov [Fri, 27 Jan 2017 14:54:01 +0000 (15:54 +0100)]
Merge pull request #3860 from masterleinad/fix_parpack_test
Update test to the new MFOperators interface
Daniel Arndt [Fri, 27 Jan 2017 14:49:59 +0000 (15:49 +0100)]
Update test to the new MFOperators interface
Martin Kronbichler [Fri, 27 Jan 2017 14:16:01 +0000 (15:16 +0100)]
Merge pull request #3857 from kalj/templatize-interface-operator
Templatize interface operator wrt vector type
Wolfgang Bangerth [Fri, 27 Jan 2017 14:14:40 +0000 (07:14 -0700)]
Merge pull request #3858 from davydden/spherical_formulas
explicitly mention Spherical-Cartesian transformation
Karl Ljungkvist [Fri, 27 Jan 2017 13:08:17 +0000 (14:08 +0100)]
templatize interface operator wrt vector type
This allows MGInterfaceOperator to work with all valid vector types, and
not just LinearAlgebra::distributed::Vector
Denis Davydov [Fri, 27 Jan 2017 12:16:12 +0000 (13:16 +0100)]
explicitly mention Spherical-Cartesian transformation
Martin Kronbichler [Fri, 27 Jan 2017 08:22:27 +0000 (09:22 +0100)]
Merge pull request #3855 from kronbichler/master
Augment list of incompatible changes for parallel vector
Martin Kronbichler [Fri, 27 Jan 2017 07:20:07 +0000 (08:20 +0100)]
Merge pull request #3849 from tjhei/mg_transfer_followup
require trilinos for MGTransferPrebuilt tests
Martin Kronbichler [Fri, 27 Jan 2017 07:18:03 +0000 (08:18 +0100)]
Augment list of incompatible changes for parallel vector
Wolfgang Bangerth [Thu, 26 Jan 2017 17:22:55 +0000 (10:22 -0700)]
Merge pull request #3854 from kronbichler/fix_compile_no_cxx11
Make compile without C++11 but Trilinos
Martin Kronbichler [Thu, 26 Jan 2017 16:06:39 +0000 (17:06 +0100)]
Make compile without C++11 but Trilinos
Bruno Turcksin [Wed, 25 Jan 2017 22:41:05 +0000 (17:41 -0500)]
Merge pull request #3848 from bangerth/doc-update-1
Update the discussion of generalized support points.
Timo Heister [Wed, 25 Jan 2017 22:37:17 +0000 (17:37 -0500)]
Merge pull request #3852 from jppelteret/fix_tests_02
Force step-44 based tests in physics to use C++11
Jean-Paul Pelteret [Wed, 25 Jan 2017 21:12:07 +0000 (22:12 +0100)]
Force step-44 based tests in physics to use C++11.
Timo Heister [Wed, 25 Jan 2017 12:11:49 +0000 (07:11 -0500)]
Merge pull request #3851 from masterleinad/fix_merge_empty
Fix ConstraintMatrix::merge for empty objects
Daniel Arndt [Wed, 25 Jan 2017 10:40:37 +0000 (11:40 +0100)]
Fix ConstraintMatrix::merge for empty objects
Timo Heister [Wed, 25 Jan 2017 01:32:35 +0000 (20:32 -0500)]
require trilinos for MGTransferPrebuilt
Wolfgang Bangerth [Wed, 25 Jan 2017 00:05:11 +0000 (17:05 -0700)]
Extend documentation.
Wolfgang Bangerth [Tue, 24 Jan 2017 23:56:35 +0000 (16:56 -0700)]
Provide a reference to the glossary.
Wolfgang Bangerth [Tue, 24 Jan 2017 23:56:07 +0000 (16:56 -0700)]
Update the discussion of generalized support points.
Wolfgang Bangerth [Tue, 24 Jan 2017 22:45:17 +0000 (15:45 -0700)]
Merge pull request #3847 from masterleinad/fix_laplcian_02
Fix integrators/laplacian_02 for 64-bit indices
Daniel Arndt [Tue, 24 Jan 2017 21:52:26 +0000 (22:52 +0100)]
Fix integrators/laplacian_02 for 64-bit indices
Daniel Arndt [Tue, 24 Jan 2017 10:03:06 +0000 (11:03 +0100)]
Merge pull request #3789 from masterleinad/merge_constraints
Allow merging of ConstraintMatrices for different local_lines
Martin Kronbichler [Tue, 24 Jan 2017 07:12:40 +0000 (08:12 +0100)]
Merge pull request #3828 from tjhei/mg_transfer_prebuilt_no_trilinos_bug
disable MGTransferPrebuilt without Trilinos
Martin Kronbichler [Tue, 24 Jan 2017 06:58:38 +0000 (07:58 +0100)]
Merge pull request #3842 from jppelteret/fix_PR_1783_01
Various fixes for PR 1783
David Wells [Mon, 23 Jan 2017 23:06:12 +0000 (18:06 -0500)]
Merge pull request #3840 from tjhei/const_time_stepping
add missing const to arguments in time_stepping.h
Jean-Paul Pelteret [Mon, 23 Jan 2017 20:59:04 +0000 (21:59 +0100)]
Amend Trilinos-related classes for when using non-MPI Trilinos build
Fixes #3841
Jean-Paul Pelteret [Mon, 23 Jan 2017 19:17:10 +0000 (20:17 +0100)]
Remove C++11 features not supported by GCC 4.6.
Fixes #3839
Timo Heister [Mon, 23 Jan 2017 17:45:21 +0000 (12:45 -0500)]
add missing const to arguments in time_stepping.h
Timo Heister [Mon, 23 Jan 2017 15:13:59 +0000 (10:13 -0500)]
Merge pull request #3838 from kronbichler/avoid_deprecated_partitioner_functions
Avoid using deprecated functions of Utilities::MPI::Partitioner
Timo Heister [Mon, 23 Jan 2017 15:13:15 +0000 (10:13 -0500)]
Merge pull request #3837 from drwells/fix-doxygen-table
Fix the default doxygen table colors.
Martin Kronbichler [Mon, 23 Jan 2017 13:28:52 +0000 (14:28 +0100)]
Avoid using deprecated functions of Utilities::MPI::Partitioner
David Wells [Mon, 23 Jan 2017 12:40:26 +0000 (07:40 -0500)]
Fix the default doxygen table colors.
Martin Kronbichler [Mon, 23 Jan 2017 07:55:46 +0000 (08:55 +0100)]
Merge pull request #3835 from tjhei/msvc_fixes
MSVC fixes
Timo Heister [Mon, 23 Jan 2017 03:13:01 +0000 (22:13 -0500)]
MSVC fixes
- enable compilation of bundled boost with MSVC 2017rc1
- remove static_assert that gets wrongly triggered during normal compilation
Timo Heister [Mon, 23 Jan 2017 02:50:22 +0000 (21:50 -0500)]
Merge pull request #3834 from tjhei/interpolate_const
Fix constness of argument in FE::interpolate()
Timo Heister [Mon, 23 Jan 2017 01:38:35 +0000 (20:38 -0500)]
Fix constness of argument in FE::interpolate()
Reported by MSVC.
Daniel Arndt [Sun, 22 Jan 2017 20:41:46 +0000 (21:41 +0100)]
Merge pull request #3833 from drwells/project-common-vector-type
Move the matrix-free project copy operation up the call stack.
Martin Kronbichler [Sun, 22 Jan 2017 19:45:42 +0000 (20:45 +0100)]
Merge pull request #3830 from jppelteret/fix_test_in_3826
Amend test introduced in #3826
David Wells [Sun, 22 Jan 2017 17:04:16 +0000 (12:04 -0500)]
Move the matrix-free project copy operation up the call stack.
The matrix-free projection solves the linear system with a specific vector type
and then copies the result into the given one. This commit moves that copy
operation into a separate function so that the actual matrix free function no
longer depends on the type of the input vector. Put another way: we no longer
have to instantiate the matrix free projections for all possible vector types.
Before moving the copy operation:
File name Memory, MB Time, s
source/numerics/vector_tools_project.cc 2774 76
source/numerics/vector_tools_project_inst2.cc 2829 81
source/numerics/vector_tools_project_inst3.cc 2823 86
After:
File name Memory, MB Time, s
source/numerics/vector_tools_project.cc 1855 47
source/numerics/vector_tools_project_inst2.cc 1995 51
source/numerics/vector_tools_project_inst3.cc 2183 55
David Wells [Sun, 22 Jan 2017 16:58:42 +0000 (11:58 -0500)]
Remove an unused local variable.
Daniel Arndt [Sun, 22 Jan 2017 16:53:59 +0000 (17:53 +0100)]
Merge pull request #3831 from drwells/project-codim-separate-file
Give the codim > 0 VectorTools::project functions their own file.
Jean-Paul Pelteret [Sun, 22 Jan 2017 07:00:05 +0000 (08:00 +0100)]
Amend test introduced in #3826
This commit attends to the comments made by @tjhei and @drwells after
#3826 was merged.
David Wells [Sun, 22 Jan 2017 05:41:09 +0000 (00:41 -0500)]
Give the codim > 0 VectorTools::project functions their own file.
Since these functions don't go into the MatrixFree code we can split them off
with a very small amount of overhead for the additional parallelization. This
lowers the required memory for the last projection instantiation by about a
gigabyte.
Before:
File name Memory, MB Time, s
source/numerics/vector_tools_project.cc 2775 79
source/numerics/vector_tools_project_inst2.cc 2882 92
source/numerics/vector_tools_project_inst3.cc 3948 102
After:
File name Memory, MB Time, s
source/numerics/vector_tools_project_codim.cc 1843 28
source/numerics/vector_tools_project.cc 2774 76
source/numerics/vector_tools_project_inst2.cc 2829 81
source/numerics/vector_tools_project_inst3.cc 2823 86
Martin Kronbichler [Sun, 22 Jan 2017 08:23:29 +0000 (09:23 +0100)]
Merge pull request #3829 from drwells/project-hp-separate-file
Give the hp version of VectorTools::project its own file.
David Wells [Sun, 22 Jan 2017 00:49:53 +0000 (19:49 -0500)]
Give the hp version of VectorTools::project its own file.
This new file takes about 17 seconds to compile and improves the performance of
the non-hp (i.e., the matrix free) projections a little bit.
before splitting out the hp files:
File name Memory, MB Time, s
source/numerics/vector_tools_project.cc 2860 81
source/numerics/vector_tools_project_inst2.cc 2970 100
source/numerics/vector_tools_project_inst3.cc 3956 114
After:
File name Memory, MB Time, s
source/numerics/vector_tools_project.cc 2775 79
source/numerics/vector_tools_project_inst2.cc 2882 92
source/numerics/vector_tools_project_inst3.cc 3948 102
These numbers come from setting the C++ compiler as a shell script containing
/usr/bin/time -v g++ ${1+"$@"}
and then parsing the output.
Timo Heister [Wed, 18 Jan 2017 22:56:52 +0000 (17:56 -0500)]
disable MGTransferPrebuilt without Trilinos
Do not allow MGTransferPrebuilt to work with distributed::Vector if we
don't have Trilinos. This used to compile and crash at runtime because
we wrongly created serial matrices.
This also fixes the tests/multigrid/transfer_prebuilt_04 error message:
An error occurred in line <361> of file
</mnt/data/testsuite/dealii/source/lac/dynamic_sparsity_pattern.cc> in
function
bool
dealii::DynamicSparsityPattern::exists(dealii::DynamicSparsityPattern::size_type,
dealii::DynamicSparsityPattern::size_type) const
The violated condition was:
rowset.size()==0 || rowset.is_element(i)
Jean-Paul Pelteret [Sat, 21 Jan 2017 17:56:21 +0000 (18:56 +0100)]
Merge pull request #3826 from bangerth/fix-trilinos-vector-reference
Fix the return type of Trilinos's VectorReference::operator=.
Wolfgang Bangerth [Sat, 21 Jan 2017 15:50:45 +0000 (08:50 -0700)]
Add test.
Wolfgang Bangerth [Sat, 21 Jan 2017 15:50:23 +0000 (08:50 -0700)]
Fix the return type of Trilinos's VectorReference::operator=.
This was non-intuitive, though it did not make a difference given the particular
semantics of vector references. Fix it anyway.
Wolfgang Bangerth [Fri, 20 Jan 2017 15:06:34 +0000 (08:06 -0700)]
Merge pull request #1783 from jppelteret/fix_trilinos_lin_ops
Fix Trilinos preconditioners and solvers for use in LinearOperator functions
Wolfgang Bangerth [Fri, 20 Jan 2017 14:41:07 +0000 (07:41 -0700)]
Merge pull request #3808 from bangerth/be-more-concise-with-update-flags
Better track which UpdateFlags are required for which operation.
Wolfgang Bangerth [Mon, 16 Jan 2017 19:09:59 +0000 (12:09 -0700)]
Add a changelog entry.
Wolfgang Bangerth [Fri, 20 Jan 2017 14:39:28 +0000 (07:39 -0700)]
Merge pull request #3813 from bangerth/doc-update-81
Better document FE_PolyTensor.
Wolfgang Bangerth [Mon, 16 Jan 2017 22:40:23 +0000 (15:40 -0700)]
Better document FE_PolyTensor.
Jean-Paul Pelteret [Mon, 16 Jan 2017 20:57:32 +0000 (21:57 +0100)]
Extensions for Trilinos support in LinearOperators.
Trilinos sparse matrices and preconditions can now we wrapped as
LinearOperators, and can be used (nearly) as naturally as the deal.II
linear algebra classes could. This means that they can be used as a
transpose_operator(), inverse_operator() and within the
schur_complement().
The core support for block operations is also there, but has not been
thoroughly tested.
Bruno Turcksin [Thu, 19 Jan 2017 21:58:38 +0000 (16:58 -0500)]
Merge pull request #3823 from bangerth/extend-signaling-nans
Implement numbers::signaling_nan() also for Point<dim>.
Wolfgang Bangerth [Thu, 19 Jan 2017 17:07:38 +0000 (10:07 -0700)]
Merge pull request #3824 from tjhei/fix_3820
fix PR 3820
Timo Heister [Thu, 19 Jan 2017 16:01:07 +0000 (11:01 -0500)]
fix PR 3820
Wolfgang Bangerth [Thu, 19 Jan 2017 01:15:16 +0000 (18:15 -0700)]
Add changelog entry.
Wolfgang Bangerth [Thu, 19 Jan 2017 01:15:03 +0000 (18:15 -0700)]
Add a test.
Wolfgang Bangerth [Thu, 19 Jan 2017 01:12:05 +0000 (18:12 -0700)]
Implement numbers::signaling_nan() also for Point<dim>.
Wolfgang Bangerth [Wed, 18 Jan 2017 20:14:00 +0000 (13:14 -0700)]
Merge pull request #3819 from danshapero/full_matrix_move
move operations for FullMatrix
Timo Heister [Wed, 18 Jan 2017 17:23:41 +0000 (12:23 -0500)]
Merge pull request #3822 from bangerth/minor-cleanups
Minor cleanups to the implementation of FETools::compute_node_matrix().
Wolfgang Bangerth [Wed, 18 Jan 2017 14:38:00 +0000 (07:38 -0700)]
Minor cleanups to the implementation of FETools::compute_node_matrix().
Daniel Arndt [Wed, 18 Jan 2017 10:28:28 +0000 (11:28 +0100)]
Allow merging with different lines only if explicitly requested
David Wells [Wed, 18 Jan 2017 00:44:09 +0000 (19:44 -0500)]
Merge pull request #3820 from tjhei/split_instantiations
split instantiations
danshapero [Tue, 17 Jan 2017 07:11:35 +0000 (23:11 -0800)]
Test for FullMatrix move operations
Timo Heister [Tue, 17 Jan 2017 22:28:07 +0000 (17:28 -0500)]
split instantiations
This splits some large files to compile with <2.5GB RAM with gcc again.
gcc 4.8.4 and using ulimit -v
2500000 except for the linking step.
Timo Heister [Tue, 17 Jan 2017 17:52:29 +0000 (12:52 -0500)]
Merge pull request #3116 from vastronzl/step-57
Introduce step-57: Navier Stokes with nonlinear solver
Timo Heister [Tue, 17 Jan 2017 16:04:55 +0000 (11:04 -0500)]
address some left-over comments
Liang Zhao [Mon, 8 Aug 2016 14:56:34 +0000 (10:56 -0400)]
Introduce step-57
Martin Kronbichler [Tue, 17 Jan 2017 17:43:22 +0000 (18:43 +0100)]
Merge pull request #3814 from bangerth/return-matrix
Let FETools::compute_node_matrix() return its result, rather than take it by reference
Daniel Arndt [Tue, 17 Jan 2017 17:01:50 +0000 (18:01 +0100)]
Extend documentation
Martin Kronbichler [Tue, 17 Jan 2017 16:22:38 +0000 (17:22 +0100)]
Merge pull request #3818 from tjhei/doxygen_fixes
doxygen fixes
Timo Heister [Tue, 17 Jan 2017 16:19:27 +0000 (11:19 -0500)]
doxygen fixes
David Wells [Tue, 17 Jan 2017 13:36:04 +0000 (08:36 -0500)]
Merge pull request #3816 from stmcgovern/typo_fix
step-40 typos
Sean McGovern [Tue, 17 Jan 2017 12:19:39 +0000 (13:19 +0100)]
step-40 typos
Jean-Paul Pelteret [Mon, 9 Jan 2017 10:14:39 +0000 (11:14 +0100)]
Make Trilinos Solver classes compatible with LinearOperators.
This patch adds some core functionality to the Trilinos Solver that is
later necessary for them to be compatible with LinearOperators.
Specifically, this is two additional solve() functions that accept
Trilinos EpetraOperators as the input Matrix (with one accomodating
wrapped Trilinos preconditioners, and the other another EpetraOperator
as a preconditioner.).
Jean-Paul Pelteret [Mon, 9 Jan 2017 10:11:58 +0000 (11:11 +0100)]
Make Trilinos Preconditioner classes compatible with LinearOperators.
This patch adds some core functionality to the Trilinos Preconditioner
that is later necessary for them to be compatible with LinearOperators.
Jean-Paul Pelteret [Mon, 16 Jan 2017 20:54:21 +0000 (21:54 +0100)]
LinearOperators now derive from an arbitrary Payload class.
This allows their functionality to be extended to support algebra class
other than deal.II's native classes.
danshapero [Tue, 17 Jan 2017 07:09:45 +0000 (23:09 -0800)]
Eliminating unnecessary copy operations from FullMatrix
The copy constructor and copy assignment operator for FullMatrix are
unnecessary, since FullMatrix inherits from Table<2>.
Wolfgang Bangerth [Mon, 16 Jan 2017 23:00:34 +0000 (16:00 -0700)]
Re-add the old function signature, mark as deprecated.
Wolfgang Bangerth [Mon, 16 Jan 2017 22:45:18 +0000 (15:45 -0700)]
Add a changelog entry.
Wolfgang Bangerth [Mon, 16 Jan 2017 22:38:07 +0000 (15:38 -0700)]
Let FETools::compute_node_matrix() return its result, rather than take it by reference.
Martin Kronbichler [Mon, 16 Jan 2017 21:43:12 +0000 (22:43 +0100)]
Merge pull request #3811 from drwells/symmetric-tensor-default-methods
Make SymmetricTensor trivially copyable.
David Wells [Mon, 16 Jan 2017 21:31:51 +0000 (16:31 -0500)]
Merge pull request #3809 from bangerth/document_FETools_compute_node_matrix
Document FETools::compute_node_matrix.
Wolfgang Bangerth [Mon, 16 Jan 2017 19:36:35 +0000 (12:36 -0700)]
Document FETools::compute_node_matrix.
David Wells [Mon, 16 Jan 2017 04:44:42 +0000 (23:44 -0500)]
Make SymmetricTensor trivially copyable.
This is a follow up to
0004f1cbcd2 which made Tensors trivially copyable.
Wolfgang Bangerth [Mon, 16 Jan 2017 19:06:18 +0000 (12:06 -0700)]
Better track which UpdateFlags are required for which operation.
Denis Davydov [Mon, 16 Jan 2017 14:13:27 +0000 (15:13 +0100)]
Merge pull request #3806 from davydden/pum_reference
fe_enriched: update bibtex reference
Denis Davydov [Mon, 16 Jan 2017 08:51:56 +0000 (09:51 +0100)]
fe_enriched: update bibtex reference
Martin Kronbichler [Mon, 16 Jan 2017 10:45:12 +0000 (11:45 +0100)]
Merge pull request #3805 from drwells/tensor-default-methods
Make Tensors trivially copyable.
David Wells [Sun, 15 Jan 2017 23:08:07 +0000 (18:08 -0500)]
Make Tensors trivially copyable.
The implicitly defined copy constructors and assignment operators are no
different from the ones implemented prior to this patch, so we can clean things
up a bit by just using the defaults here.
This has the nice effect of making the class "trivially copyable", which means
compilers may optimize things by using memcpy to copy things instead of the copy
constructor.
David Wells [Sun, 15 Jan 2017 21:36:27 +0000 (16:36 -0500)]
Merge pull request #3804 from Rombur/fix_shared_ptr
Fix a test.
Bruno Turcksin [Sun, 15 Jan 2017 20:45:38 +0000 (15:45 -0500)]
Fix a test.
Jean-Paul Pelteret [Sun, 15 Jan 2017 09:41:57 +0000 (10:41 +0100)]
Merge pull request #3803 from tamiko/a_default_parameter
Bugfix: Add forgotten default parameter to function
Martin Kronbichler [Sun, 15 Jan 2017 08:19:58 +0000 (09:19 +0100)]
Merge pull request #3801 from drwells/remove-Reference-friendship
Get rid of some unused friendship declarations.
Martin Kronbichler [Sun, 15 Jan 2017 08:18:48 +0000 (09:18 +0100)]
Merge pull request #3799 from kronbichler/master
Cleanup code in vectorization
In the beginning the Universe was created. This has made a lot of
people very angry and has been widely regarded as a bad move.
Douglas Adams
Typeset in
Trocchi and Trocchi Bold Sans Serif.