ForkMaster/netgen
1
0
Fork 0
mirror of https://github.com/NGSolve/netgen.git synced 2024-11-13 17:48:34 +05:00
Code Issues Packages Projects Releases Wiki Activity
1,846 Commits 10 Branches 43 Tags 18 MiB
da5c9723d1
Commit Graph

173 Commits

Author SHA1 Message Date
Matthias Hochsteger
9247c89fd7 add missing function HSum(a,b) for SIMD<double,1> 2017-04-20 11:41:23 +02:00
Matthias Hochsteger
643c89538d Add support for AVX512 
Restructure mysimd.hpp and add support for AVX512. Changes include

  - Move mysimd.hpp to ngsimd.hpp
  - Expose ngsimd.hpp to NGSolve
  - New namespace ngsimd
  - Second template parameter (width) for SIMD class, default to the
    largest width available
  - Avoid raw avx register types in the interface, use SIMD<> instead
 2017-04-19 18:02:27 +02:00
Joachim Schöberl
4382fc1168 fix hashtable 2017-04-11 10:04:33 +02:00
Joachim Schöberl
e6b853e995 hashtables roundup to power of 2, optimize bisect 2017-04-11 09:01:36 +02:00
Lukas Kogler
d24b996c8f Merge branch 'master' into par_fixes 2017-03-22 14:03:11 +01:00
Matthias Hochsteger
079406f615 Include <functional> (gcc7 was complaining here) 2017-03-17 21:09:27 +01:00
lkogler
383ced496b Changed MyMPI_RecvCmd and MyMPI_SendCmd. Disabled USE_BUFFERS for now. 2017-03-06 14:32:20 +01:00
Joachim Schöberl
90d4c94028 make Topology a member of Mesh (rather then pointer) to reduce number of pointer dereferencing steps. This required move operators for Array and Table. 2017-02-25 19:48:37 +01:00
Joachim Schöberl
c143aafaea Array resize checks for trivially copyable 2017-02-24 22:29:57 +01:00
Matthias
f674d5a20a Assume that malloc returns memory that is 16 byte-aligned 
This fixes an issue with inconsistent memory allocation/deallocation of
MeshPoint in Python.
 2017-02-07 10:26:31 +01:00
Matthias Hochsteger
4cff6b546a install all headers to INCDIR 2017-01-16 14:59:57 +01:00
Joachim Schöberl
59dc0b6c6a Array::Append does not return size anymore ( will be changed to last ) 2016-12-11 18:22:07 +01:00
Matthias Hochsteger
ee4d5e89d8 remove unnecessary files (mostly unsupported build systems) 2016-11-23 18:07:14 +01:00
Christopher Lackner
0a9abc1acc array iterator to python 2016-11-15 18:12:51 +01:00
Matthias Hochsteger
5b253c76e0 Add a guard to avoid multiply defined AVX operators on Windows 2016-11-10 15:36:47 +01:00
Lukas Kogler
37e0e17919 removed some comments. formatting 2016-11-08 12:57:59 +01:00
Lukas Kogler
49aa2861d6 changed buffer size in send/recv cmd (for sending entire python files) 2016-11-07 17:48:23 +01:00
Matthias Hochsteger
399d370463 fix some typos and add placement new operator to AlignedAlloc 2016-11-04 12:48:23 +01:00
Matthias Hochsteger
96c1235184 switch from boost::python to pybind11 2016-11-04 12:14:52 +01:00
Joachim Schöberl
3fcb7d13d5 dummy aligned-alloc without AVX 2016-10-30 15:15:16 +01:00
Joachim Schöberl
3a631f10ca solution visualization using AVX 2016-10-30 15:01:52 +01:00
Joachim Schöberl
cdcd868712 size_t 2016-10-25 23:34:06 +02:00
Joachim Schöberl
e009a8b687 started task-manager in netgen (topology) 2016-08-18 00:49:47 +02:00
Joachim Schöberl
5448157842 use SIMD also for AVX (not only AVX2) 2016-08-07 19:13:36 +02:00
Alexander Schlüter
d282a5914d fix compiler errors in mpi build 2016-07-15 15:26:25 +02:00
Joachim Schöberl
0a9adc91e9 default simd ctr 2016-07-07 19:54:47 +02:00
Joachim Schöberl
59bc017c40 bring std-operators to namespace 2016-07-07 19:49:02 +02:00
Joachim Schöberl
8414cb2d60 using simds for element-trafo, added type-template to many classes 2016-07-06 17:12:57 +01:00
Joachim Schoeberl
ee6fbe65e7 parallel refinement 2016-02-28 06:05:19 +01:00
Joachim Schöberl
1ac9c02f5b lexicographic ordering for uni-form mesh refinement 2016-02-26 20:29:14 +01:00
Matthias
8a3e29e624 get rid of pthread 2016-02-23 20:23:48 +01:00
Matthias Hochsteger
402e85cc17 bugfix due to capturing of thread index by reference 2016-02-23 11:08:20 +01:00
Matthias Hochsteger
1b4f596446 get rid of OpenMP 2016-02-22 18:44:00 +01:00
Matthias
11ac6c704a fixes for Intel Compiler on Windows 2015-11-02 08:06:18 -08:00
Matthias Hochsteger
8ac965d707 everything in one big dll on Windows 2015-10-19 10:08:30 +02:00
Joachim Schoeberl
caa7295990 archive with Input/Ouput variable 2015-07-23 14:10:50 +02:00
Matthias Hochsteger
bd18af5676 cmake install components 2015-04-09 12:27:59 +02:00
Joachim Schoeberl
631b519676 archive array-functions 2015-03-23 14:51:05 +01:00
Matthias Hochsteger
b870bcee0f CMake support 2015-03-06 14:59:14 +01:00
Joachim Schoeberl
257e76c083 wrappers for netgen-dll 2015-01-27 15:56:32 +00:00
Gerhard Kitzler
785214bbab added ngpython.hpp 2015-01-27 14:29:51 +00:00
Joachim Schoeberl
69429c6e09 occ meshing, lock in block-allocator 2015-01-20 17:41:16 +00:00
Joachim Schoeberl
e920593821 add size-t 2015-01-08 12:47:27 +00:00
Joachim Schoeberl
8e8e390f2e 2d Delaunay, array-iterators 2014-12-02 13:23:36 +00:00
Joachim Schoeberl
36803ec256 mesh interface, deallocation 2014-10-19 13:53:57 +00:00
Christoph Wintersteiger
4bef2d7617 ModuleScope for Python modules added 2014-10-08 15:48:01 +00:00
Joachim Schoeberl
11c641b038 ngpython 2014-10-08 14:09:03 +00:00
Joachim Schoeberl
f96adac6fe autoptr and parallel 2014-09-10 16:15:03 +00:00
Joachim Schoeberl
e9657ab1bb archive 2014-04-28 06:54:06 +00:00
Joachim Schoeberl
d0f82c3993 archive 2014-03-04 12:27:15 +00:00
Joachim Schoeberl
3a82cf2258 archive 2014-02-23 19:31:40 +00:00
Gerhard Kitzler
18ba82517f DLL_HEADER 2013-04-05 13:08:07 +00:00
Joachim Schoeberl
8d7debaf9b indirect arrays 2013-04-02 20:26:21 +00:00
Joachim Schoeberl
7d4c806012 extern printdots 2013-03-05 10:25:17 +00:00
Joachim Schoeberl
b225adb2cd compressed mesh files (zlib) 2013-02-27 17:29:12 +00:00
Joachim Schoeberl
7aa0128d6f add zlib / gzstream 2013-02-26 12:59:12 +00:00
Gerhard Kitzler
6c48d43c27 DLL_HEADER 2012-10-29 17:45:15 +00:00
Joachim Schoeberl
5dde1907d6 profiler 2012-10-27 11:47:21 +00:00
Joachim Schoeberl
c7fb4b676f improvements for STL meshing 2012-10-22 13:13:57 +00:00
Joachim Schoeberl
8ef8c2e975 improved exchangetable 2012-09-03 09:49:18 +00:00
Joachim Schoeberl
7233ecdcd5 mpi - table exchange improved 2012-08-30 19:24:20 +00:00
Joachim Schoeberl
6a1e8f7e97 parallel topology + curved elements 2012-08-20 14:10:23 +00:00
Joachim Schoeberl
1d163e69dc get arguments 2012-07-06 07:49:05 +00:00
Joachim Schoeberl
a7d1c6c657 sequential fixes 2012-07-05 13:02:01 +00:00
Joachim Schoeberl
312dcb3e31 parallel 2012-06-18 18:40:05 +00:00
Joachim Schoeberl
eade1f6b2d mpi 2012-06-16 16:58:46 +00:00
Joachim Schoeberl
05b0abfce1 mpi 2012-06-16 16:58:28 +00:00
Joachim Schoeberl
51084912d9 vampir 2012-06-16 15:22:46 +00:00
Joachim Schoeberl
2c3d08ec2a mpi - table exchange 2012-06-16 12:03:36 +00:00
Joachim Schoeberl
27f8e452fd parallel fixes 2012-06-13 09:07:11 +00:00
Joachim Schoeberl
6741b74a54 fixes by matthias 2012-05-24 11:12:37 +00:00
Joachim Schoeberl
cfc7c46df0 tiny parallel polish 2011-11-03 11:03:11 +00:00
Joachim Schoeberl
fec9740f4c metis 5 adaption 2011-11-01 12:54:07 +00:00
Joachim Schoeberl
f5f099d868 export mpi variables 2011-08-31 17:31:59 +00:00
Joachim Schoeberl
577ecd6f47 curved elements support MPI and multithreading 2011-08-11 22:28:22 +00:00
Joachim Schoeberl
8a450b506b remove debugging output 2011-08-08 19:48:12 +00:00
Joachim Schoeberl
b90dbcfbe5 needed for curved elements in parallel 2011-08-08 13:34:08 +00:00
Joachim Schoeberl
a582fc6231 on the way to thread-save meshing 2011-07-25 08:40:23 +00:00
Joachim Schoeberl
342983ee75 parallel topology simplification 2011-07-14 21:36:19 +00:00
Joachim Schoeberl
3aeb3d79b9 improved parallel mesh loading 2011-07-13 18:32:11 +00:00
Joachim Schoeberl
70292000c3 interface 2011-07-13 18:27:17 +00:00
Joachim Schoeberl
28f59f7bcf parallel communication tags 2011-07-06 21:08:58 +00:00
Joachim Schoeberl
2d15141187 parallel mesh 2011-07-04 12:29:18 +00:00
Joachim Schoeberl
a64aa8226e parallelization 2011-06-30 12:49:38 +00:00
Joachim Schoeberl
660f31d7cd parallelization 2011-06-26 07:35:08 +00:00
Joachim Schoeberl
580f4b9f52 parallel ngs 2011-06-16 17:55:08 +00:00
Joachim Schoeberl
3d2f7d7457 dll_headers for geometry dll (only 2D yet) 2011-03-03 21:42:20 +00:00
Joachim Schoeberl
5204be0657 spline files restructured 2011-02-28 12:59:27 +00:00
Joachim Schoeberl
b0f637862b parallel 2011-02-18 19:58:17 +00:00
Joachim Schoeberl
846542c45c parallel programming 2011-02-14 12:27:18 +00:00
Joachim Schoeberl
b95acb082c geometry restructuring 2011-01-10 20:18:01 +00:00
Joachim Schoeberl
75a6623419 2d meshing improvements 2010-07-19 12:00:52 +00:00
Joachim Schoeberl
768ef87cac suggestions by E.A. (salome) 2010-05-18 11:28:15 +00:00
Joachim Schoeberl
20c2f2757e quicksort 2009-10-28 00:22:10 +00:00
Joachim Schoeberl
94a06d8cae quicksort for arrays 2009-10-28 00:04:19 +00:00
Joachim Schoeberl
6f39164242 reduced memory requirement for the setup of topology faces 2009-10-27 18:14:45 +00:00
Joachim Schoeberl
41de0eaa1a remove moveable mem 2009-08-25 14:01:10 +00:00
Joachim Schoeberl
670509b52b baseclass NetgenGeometry 2009-08-25 14:00:20 +00:00
Joachim Schoeberl
f14a04a837 basegeometry, disable moveablemem 2009-08-24 00:03:40 +00:00
Joachim Schoeberl
f3043d2df9 improvement in visualization 2009-07-20 08:36:36 +00:00