]>
https://gitweb.dealii.org/ - dealii.git/log
David Wells [Sat, 6 Apr 2024 17:15:47 +0000 (13:15 -0400)]
execute_refinement_isotropic(): Add a loop.
David Wells [Sat, 6 Apr 2024 17:15:17 +0000 (13:15 -0400)]
execute_refinement_isotropic(): Clean up some comments.
David Wells [Sat, 6 Apr 2024 15:42:29 +0000 (11:42 -0400)]
execute_refinement_isotropic(): Generalize 2d child line assignment.
Quadrilateral cells in mixed meshes may have lines with reversed orientations.
We can make that case work correctly by simply generalizing the code we already
have for triangles.
David Wells [Sat, 6 Apr 2024 14:37:37 +0000 (10:37 -0400)]
Split a very long string.
Matthias Maier [Sat, 6 Apr 2024 05:35:00 +0000 (05:35 +0000)]
Merge pull request #16858 from tamiko/fix_cgal_test_fpe
Test: cgal/cgal_mesh_criteria: disable floating point exceptions
Wolfgang Bangerth [Sat, 6 Apr 2024 04:36:23 +0000 (22:36 -0600)]
Merge pull request #16840 from tamiko/vectorization_type_trait
Vectorization type trait
Matthias Maier [Sat, 6 Apr 2024 03:34:16 +0000 (03:34 +0000)]
Merge pull request #16859 from peterrum/remove_debug_output
Remove debug output
Matthias Maier [Fri, 5 Apr 2024 19:57:13 +0000 (14:57 -0500)]
Test: cgal/cgal_mesh_criteria: disable floating point exceptions
We trigger a (deliberate) floating point exception in the CGAL (header)
library:
```
// The multiplications could produce some NaN, with 0 * inf. Replacing it with inf is safe.
// min(x,y) (the order is essential) returns its second argument when the first is NaN.
// An IEEE 754-2019 maximum could help.
__m128d big = IA::largest().simd();
-> __m128d x1 = _mm_mul_pd(aa,bz); // {-ai*bi,as*bs}
//x1 = _mm_min_pd(x1,big); // no NaN
__m128d x2 = _mm_mul_pd(aa,c); // {-ai*bs,as*bi}
x2 = _mm_min_pd(x2,big); // no NaN
__m128d x3 = _mm_mul_pd(ap,bz); // {-as*bi,ai*bs}
//x3 = _mm_min_pd(x3,big); // no NaN
__m128d x4 = _mm_mul_pd(ap,c); // {-as*bs,ai*bi}
x4 = _mm_min_pd(x4,big); // no NaN
```
Thus, disable floating point exceptions for the test.
Peter Munch [Fri, 5 Apr 2024 20:22:55 +0000 (22:22 +0200)]
Remove debug output
Wolfgang Bangerth [Fri, 5 Apr 2024 17:37:00 +0000 (11:37 -0600)]
Merge pull request #16853 from kronbichler/avoid_allocations_2
Add Quadrature::initialize() function taking ArrayView
Wolfgang Bangerth [Fri, 5 Apr 2024 17:29:29 +0000 (11:29 -0600)]
Merge pull request #16624 from peterrum/matrixcreator_fullmatrix
Enable MatrixCreator functions also for FullMatrix
Martin Kronbichler [Fri, 5 Apr 2024 14:54:37 +0000 (16:54 +0200)]
Merge both initialize functions
Wolfgang Bangerth [Fri, 5 Apr 2024 14:27:11 +0000 (08:27 -0600)]
Merge pull request #16852 from kronbichler/fix_docu
MappingQ: Remove outdated documentation
Daniel Arndt [Fri, 5 Apr 2024 13:31:49 +0000 (09:31 -0400)]
Merge pull request #16855 from kronbichler/fix_pow
Avoid std::pow in BarycentricPolynomials
Martin Kronbichler [Fri, 5 Apr 2024 10:31:30 +0000 (12:31 +0200)]
Avoid std::pow in BarycentricPolynomials
Martin Kronbichler [Fri, 5 Apr 2024 08:47:09 +0000 (10:47 +0200)]
New test case
Martin Kronbichler [Fri, 5 Apr 2024 08:30:17 +0000 (10:30 +0200)]
Add Quadrature::initialize() function taking ArrayView
Martin Kronbichler [Fri, 5 Apr 2024 08:02:14 +0000 (10:02 +0200)]
Remove outdated documentation
Martin Kronbichler [Fri, 5 Apr 2024 07:05:18 +0000 (09:05 +0200)]
Merge pull request #16849 from tamiko/fix_fe_values_tests
Test fe/fe_values_view_{07|10|13|16}: make Assert more robust
Matthias Maier [Fri, 5 Apr 2024 03:19:25 +0000 (22:19 -0500)]
Test fe/fe_values_view_{07|10|13|16}: make Assert more robust
The Assert checks that we computed with a relative error of less than
1e-12. This fails however, if the vector in question is 0. Thus, also
check for an absolute error.
Peter Munch [Sat, 10 Feb 2024 11:48:15 +0000 (12:48 +0100)]
Enable MatrixCreator functions also for FullMatrix
Martin Kronbichler [Thu, 4 Apr 2024 18:13:53 +0000 (20:13 +0200)]
Merge pull request #16847 from kronbichler/avoid_allocations
MappingQ: Favor get_vertices() over compute_mapping_support_points() if possible
Daniel Arndt [Thu, 4 Apr 2024 16:39:19 +0000 (12:39 -0400)]
Merge pull request #16830 from kronbichler/mapping_cartesian
MappingCartesian: Some performance improvements
Martin Kronbichler [Wed, 3 Apr 2024 20:46:41 +0000 (22:46 +0200)]
MappingQ: Use get_vertices() instead of compute_mapping_support_points when possible
Wolfgang Bangerth [Thu, 4 Apr 2024 05:48:35 +0000 (23:48 -0600)]
Merge pull request #16800 from kronbichler/fe_eval_docu
FEEvaluation: Modernize documentation of get_xxx/submit_xxx
Matthias Maier [Thu, 4 Apr 2024 04:38:28 +0000 (04:38 +0000)]
Merge pull request #16843 from kronbichler/improve_vector_tools_interpolate
VectorTools::interpolate: Make common case of vector problem faster
Matthias Maier [Thu, 4 Apr 2024 04:34:27 +0000 (04:34 +0000)]
Merge pull request #16839 from bangerth/tensor-product
Make tensor*tensor implementation easier to read.
Matthias Maier [Tue, 2 Apr 2024 20:55:11 +0000 (15:55 -0500)]
VectorizedArray: add constexpr boolean that records whether a type is vectorized
Matthias Maier [Tue, 2 Apr 2024 21:07:17 +0000 (16:07 -0500)]
Vectorization: move assert into constructor
Martin Kronbichler [Wed, 3 Apr 2024 18:03:54 +0000 (20:03 +0200)]
Merge pull request #16844 from kronbichler/eval_interface
Use ArrayView for interface to tensor-product point evaluator
Wolfgang Bangerth [Wed, 3 Apr 2024 17:11:43 +0000 (11:11 -0600)]
Merge pull request #16845 from kronbichler/pass_by_reference
Pass some SmartPointer variables by reference
Martin Kronbichler [Wed, 3 Apr 2024 11:39:35 +0000 (13:39 +0200)]
Use ArrayView for interface to tensor-product point evaluator
Martin Kronbichler [Wed, 3 Apr 2024 13:10:51 +0000 (15:10 +0200)]
Merge pull request #16832 from bangerth/fix-tests
Fix naming of output files.
Martin Kronbichler [Wed, 3 Apr 2024 13:10:32 +0000 (15:10 +0200)]
Merge pull request #16841 from bangerth/type-to-string
Augment documentation of Utilities::type_to_string().
Martin Kronbichler [Wed, 3 Apr 2024 10:52:07 +0000 (12:52 +0200)]
Changelog
Martin Kronbichler [Wed, 3 Apr 2024 10:50:35 +0000 (12:50 +0200)]
VectorTools::interpolate: Make common case of vector problem faster
Martin Kronbichler [Wed, 3 Apr 2024 12:01:13 +0000 (14:01 +0200)]
Pass some SmartPointer variables by reference
Martin Kronbichler [Wed, 3 Apr 2024 07:28:19 +0000 (09:28 +0200)]
Merge pull request #16837 from bangerth/vec-array-2
Provide an implementation for VectorizedArray::dot_product().
Wolfgang Bangerth [Wed, 3 Apr 2024 03:09:14 +0000 (21:09 -0600)]
Augment documentation of Utilities::type_to_string().
Rene Gassmoeller [Tue, 2 Apr 2024 18:20:10 +0000 (14:20 -0400)]
Merge pull request #16833 from kronbichler/fgmres_orth
Switch SolverFGMRES to new delayed classical Gram-Schmidt
Martin Kronbichler [Tue, 2 Apr 2024 16:37:29 +0000 (18:37 +0200)]
Improve two assertions
Wolfgang Bangerth [Tue, 2 Apr 2024 16:11:03 +0000 (10:11 -0600)]
Make tensor*tensor implementation easier to read.
Martin Kronbichler [Mon, 1 Apr 2024 19:22:43 +0000 (21:22 +0200)]
Changelog
Martin Kronbichler [Mon, 1 Apr 2024 19:20:36 +0000 (21:20 +0200)]
Test case
Martin Kronbichler [Mon, 1 Apr 2024 19:20:23 +0000 (21:20 +0200)]
MappingCartesian: Some performance improvements
Martin Kronbichler [Tue, 2 Apr 2024 15:10:12 +0000 (17:10 +0200)]
Merge pull request #16828 from drwells/qprojector-same-orientation
QProjector: internally use the combined orientation.
Wolfgang Bangerth [Tue, 2 Apr 2024 03:23:08 +0000 (21:23 -0600)]
Provide a default implementation for VectorizedArray::dot_product().
David Wells [Sun, 31 Mar 2024 20:02:27 +0000 (16:02 -0400)]
QProjector: internally use the combined orientation.
This replaces a unique indexing scheme for orientations with the standard one.
David Wells [Tue, 2 Apr 2024 13:24:31 +0000 (09:24 -0400)]
FE_BDM: un-hardcode some QProjector properties.
Martin Kronbichler [Tue, 2 Apr 2024 07:13:37 +0000 (09:13 +0200)]
Merge pull request #16836 from bangerth/vec-array
Minor updates to VectorizedArray.
Wolfgang Bangerth [Tue, 2 Apr 2024 03:11:14 +0000 (21:11 -0600)]
Minor updates to VectorizedArray.
David Wells [Tue, 2 Apr 2024 02:35:01 +0000 (22:35 -0400)]
Merge pull request #16831 from kronbichler/fix_n_vertices
Fix two places where TriaAccessor::n_vertices is called repeatedly
Martin Kronbichler [Mon, 1 Apr 2024 21:04:37 +0000 (23:04 +0200)]
Switch SolverFGMRES to new delayed classical Gram-Schmidt
Wolfgang Bangerth [Mon, 1 Apr 2024 17:55:40 +0000 (11:55 -0600)]
Fix naming of output files.
Martin Kronbichler [Mon, 1 Apr 2024 19:27:41 +0000 (21:27 +0200)]
Fix two places where CellAccessor::n_vertices is called repeatedly
Martin Kronbichler [Mon, 1 Apr 2024 16:49:47 +0000 (18:49 +0200)]
Merge pull request #16821 from drwells/qprojector-description-orientation
Deprecate some more (bool, bool, bool) orientation functions
David Wells [Mon, 1 Apr 2024 13:12:28 +0000 (09:12 -0400)]
Improve a comment.
David Wells [Sat, 30 Mar 2024 12:24:29 +0000 (08:24 -0400)]
Remove calls to said functions on the library.
David Wells [Sat, 30 Mar 2024 11:57:44 +0000 (07:57 -0400)]
Deprecate (bool, bool, bool) QProjector functions.
Daniel Arndt [Mon, 1 Apr 2024 12:33:36 +0000 (08:33 -0400)]
Merge pull request #16804 from bangerth/comsol-tri
Add a test for COMSOL .mphtxt triangle meshes.
Martin Kronbichler [Sun, 31 Mar 2024 18:10:37 +0000 (20:10 +0200)]
Merge pull request #16823 from bangerth/unalign
Revert aligning tensors
Wolfgang Bangerth [Thu, 28 Mar 2024 18:35:31 +0000 (12:35 -0600)]
Add a test for COMSOL .mphtxt triangle meshes.
Wolfgang Bangerth [Sat, 30 Mar 2024 15:04:10 +0000 (09:04 -0600)]
Revert "No longer output the size of tensors into output file of a test."
This reverts commit
a492d57657769bdc38a65c2532f1711a55abab21 .
Wolfgang Bangerth [Sat, 30 Mar 2024 15:03:43 +0000 (09:03 -0600)]
Revert "Add a changelog entry for the alignment of tensors."
This reverts commit
d14a12db1cc7b91380a143350e613442362c876e .
Wolfgang Bangerth [Sat, 30 Mar 2024 15:03:26 +0000 (09:03 -0600)]
Revert "Align Tensor<rank,dim,Number> to make vectorization possible."
This reverts commit
d6a65a9b929b1862a78de843ceaa7848e5a588e3 .
Wolfgang Bangerth [Sat, 30 Mar 2024 15:01:57 +0000 (09:01 -0600)]
Merge pull request #16820 from tamiko/fix_avx512_fpe
distributed/tria.cc: avoid FPE due to undefined behavior
Wolfgang Bangerth [Sat, 30 Mar 2024 14:59:56 +0000 (08:59 -0600)]
Merge pull request #16814 from simonsticko/codim_in_getname
Add codimension in get_name() for a few elements
Matthias Maier [Sat, 30 Mar 2024 00:49:16 +0000 (19:49 -0500)]
distributed/tria.cc: avoid FPE due to undefined behavior
The following construct
```
static_cast<types::subdomain_id>(this_sc_point[dim])
```
triggers a floating point exception when the stored double is -1.0 (at
least with gcc-11 and avx512 instructions enabled).
Judging from cppreference [1] this is undefined behavior:
"""
A prvalue of floating-point type can be converted to a prvalue of any
integer type. The fractional part is truncated, that is, the fractional
part is discarded.
If the truncated value cannot fit into the destination type, the
behavior is undefined (even when the destination type is unsigned,
modulo arithmetic does not apply).
"""
Thus, work around the issue by explicitly setting negative values to
`numbers::invalid_subdomain_id` if we encounter -1.
[1] https://en.cppreference.com/w/cpp/language/implicit_conversion
Matthias Maier [Sat, 30 Mar 2024 00:28:21 +0000 (00:28 +0000)]
Merge pull request #16816 from bangerth/fix-test
No longer output the size of tensors into output file of a test.
Matthias Maier [Sat, 30 Mar 2024 00:27:36 +0000 (00:27 +0000)]
Merge pull request #16815 from bangerth/fix-particle-size-2
Fix a second place where particles compute the size of a buffer.
Wolfgang Bangerth [Fri, 29 Mar 2024 22:30:32 +0000 (16:30 -0600)]
No longer output the size of tensors into output file of a test.
Wolfgang Bangerth [Fri, 29 Mar 2024 19:21:50 +0000 (13:21 -0600)]
Fix a second place where particles compute the size of a buffer.
Martin Kronbichler [Fri, 29 Mar 2024 19:49:19 +0000 (20:49 +0100)]
Merge pull request #16813 from bangerth/fix-particle-size
Allocate the right amount of memory when serializing particles.
Daniel Arndt [Fri, 29 Mar 2024 17:30:56 +0000 (13:30 -0400)]
Merge pull request #16788 from kronbichler/fix_asserts
FEEvaluation: Add static asserts to functions selected by enable_if
Simon Sticko [Fri, 29 Mar 2024 08:11:57 +0000 (09:11 +0100)]
Add codimension in get_name() for a few elements
Martin Kronbichler [Fri, 29 Mar 2024 16:46:53 +0000 (17:46 +0100)]
Merge pull request #16795 from Rombur/python_binding
Add three GridGenerator functions to the python wrappers
Wolfgang Bangerth [Fri, 29 Mar 2024 16:17:56 +0000 (10:17 -0600)]
Allocate the right amount of memory when serializing particles.
Daniel Arndt [Fri, 29 Mar 2024 12:20:03 +0000 (08:20 -0400)]
Merge pull request #16806 from drwells/remove-hardcoded-inverse-orientation
Remove hardcoded inverse orientation
Martin Kronbichler [Fri, 29 Mar 2024 07:06:47 +0000 (08:06 +0100)]
Merge pull request #16812 from tamiko/fix_tests_avx_05
Test grid/closest_point: guard against roundoff error
Martin Kronbichler [Fri, 29 Mar 2024 07:06:05 +0000 (08:06 +0100)]
Merge pull request #16811 from tamiko/fix_tests_avx_04
Test trilinos/precondition: Relax iteration bounds
Martin Kronbichler [Fri, 29 Mar 2024 07:05:18 +0000 (08:05 +0100)]
Merge pull request #16810 from tamiko/fix_tests_avx_03
Test grid/grid_generator_09: add another output variant
Martin Kronbichler [Fri, 29 Mar 2024 07:04:35 +0000 (08:04 +0100)]
Merge pull request #16809 from tamiko/fix_tests_avx_02
Test mappings/mapping_q_real_to_unit_internal: add avx256 output variant
Martin Kronbichler [Fri, 29 Mar 2024 07:02:56 +0000 (08:02 +0100)]
Merge pull request #16808 from tamiko/fix_tests_avx_01
Test cgal/cgal_triangulation_05: add avx output variants
Martin Kronbichler [Fri, 29 Mar 2024 07:01:58 +0000 (08:01 +0100)]
Merge pull request #16805 from simonsticko/hp_get_present_reinit_not_called
Add assert on reinit() being called before get_present_fe_values()
David Wells [Mon, 25 Mar 2024 16:26:36 +0000 (12:26 -0400)]
GridTools: remove another hard-coded inverse orientation table.
David Wells [Mon, 25 Mar 2024 16:25:35 +0000 (12:25 -0400)]
ReferenceCell: permit face index calculations for lines.
This way we can call standard_to_real_face_vertex for all supported topological
dimensions.
Matthias Maier [Fri, 29 Mar 2024 00:15:40 +0000 (19:15 -0500)]
Test grid/closest_point: guard against roundoff error
Matthias Maier [Fri, 29 Mar 2024 00:00:13 +0000 (19:00 -0500)]
Test trilinos/precondition: Relax iteration bounds
Matthias Maier [Thu, 28 Mar 2024 23:37:11 +0000 (18:37 -0500)]
Test grid/grid_generator_09: add another output variant
Matthias Maier [Thu, 28 Mar 2024 23:27:50 +0000 (18:27 -0500)]
Test mappings/mapping_q_real_to_unit_internal: add avx256 output variant
Matthias Maier [Thu, 28 Mar 2024 22:58:01 +0000 (17:58 -0500)]
Test cgal/cgal_triangulation_05: add avx output variants
The order of vertices that cgal writes out to the mesh are not stable
and depend on optimization, standard library and instruction set used.
Add output variants for avx256/avx512
David Wells [Mon, 25 Mar 2024 15:03:05 +0000 (11:03 -0400)]
Triangulation: remove hard-coded inverse orientation tables.
Peter Munch [Thu, 28 Mar 2024 20:06:43 +0000 (21:06 +0100)]
Merge pull request #16780 from drwells/consistent-periodic-orientation-encoding
Consistent periodic orientation encoding
Wolfgang Bangerth [Thu, 28 Mar 2024 19:55:52 +0000 (13:55 -0600)]
Merge pull request #16801 from kronbichler/fix_tensor_mapping
MappingQ: Do not assume contiguous storage of Tensor<2, dim>
Simon Sticko [Thu, 28 Mar 2024 08:29:10 +0000 (09:29 +0100)]
Add assert on reinit() being called before get_present_fe_values()
If you call hp::FEValues::get_present_fe_values() and forgot to call
reinit() first, you presently get a poor error message. Add an assert
that present_fe_values_index is not invalid with a better error
message.
Daniel Arndt [Thu, 28 Mar 2024 18:32:11 +0000 (14:32 -0400)]
Merge pull request #16802 from bangerth/fix-test
Fix compilation of a test.
Martin Kronbichler [Thu, 28 Mar 2024 18:07:12 +0000 (19:07 +0100)]
Merge pull request #16803 from bangerth/doc
Augment documentation of class Tensor.
Wolfgang Bangerth [Thu, 28 Mar 2024 17:52:41 +0000 (11:52 -0600)]
Merge pull request #16797 from tamiko/make_statement_more_readable
`base/tensor.h`: make a statement a bit easier to read
Martin Kronbichler [Thu, 28 Mar 2024 14:57:15 +0000 (15:57 +0100)]
MappingQ: Do not assume contiguous storage of Tensor<2, dim>
Wolfgang Bangerth [Thu, 28 Mar 2024 16:12:30 +0000 (10:12 -0600)]
Augment documentation of class Tensor.
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.