ChangeLog 14.6 KB
Newer Older
1 2 3 4 5

        * tnlBoundaryConditionsSetter was implemented 
           - it allows to set boundary conditions to given function
           - in explicit solvers, it must be called explicitly now after calling tnlExplicitUpdater

6 7 8 9 10 11 12 13 14 15 16
2015-12-25 Tomas Oberhuber  <tomas.oberhuber@fjfi.cvut.cz>
        * significant changes in grids
            - the central structure is now grid entity rather than grid itself
            - specialization of cells for cross stencils was implemented -- they store indexes of neighbhour cells
            - it currently does not work with nvcc compiler
            - grid traversers takes entity type instead of entity dimensions as a templated argument
        * mesh function was implemented
        * boundary conditions were refactored a simplified
            - there is no need to have analytic and discrete types of boundary conditions
        * tnl-quickstart does not generate correct code now

17 18 19 20 21 22 23
2012-10-05  Tomas Oberhuber  <tomas.oberhuber@fjfi.cvut.cz>
 
	* src/mesh/detail/container.h:
	  replacing Array for m_data with tnlArrayManager     
	* tests/mesh/main.cpp (main):
	  re-defining i/o files w.r.t. the project directory structure 

24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43
2012-07-09  oberhuber  <oberhuber@yvonne>

	* src/core/tnlVectorBase.h: Removed.
	* src/core/tnlArrayManager.h: New file.
	* src/core/tnlArrayManagerBase.h: New file.
	* src/core/tnlArrayManagerCUDA.h: New file.
	* src/core/tnlArrayManagerHost.h: New file.
	* tests/unit-tests/core/tnlArrayManagerTester.h: New file.
	  - adding tnlArrayManager for basic operations with arrays - allocation/deallocation, load/save from/to file
	* CMakeLists.txt:
	  - fixing problem with linking of TNL, it is build as one library 
	* src/core/CMakeLists.txt: 
	* src/core/tnlCommunicator.h: 
	* src/core/tnlObject.h (tnlObject): 
	* src/core/tnlVector.h: 
	  - refactoring tnlVector - some methods were moved to tnlArrayManager
	* src/core/tnlVectorCUDA.h: 
	* src/core/tnlVectorHost.h: 
	* TODO: 

44 45 46 47 48 49 50 51 52
2011-06-29  oberhuber  <oberhuber@yvonne>

	* cmake/AddCompilerFlag.cmake: New file.
	* cmake/MacroEnsureVersion.cmake: New file.
	* cmake/OptimizeForArchitecture.cmake: New file.
	* src/matrix/.tnlRgCSRMatrix.h.swp: New file.
	* src/matrix/tnlAdaptiveRgCSRMatrix.h: 
	 - optimizing Adaptive Rg CSR format by adding break to SpMV loop.

53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72
2011-06-15  oberhuber  <oberhuber@yvonne>

	* src/core/tnlCudaSupport.cpp: Removed.
	* CMakeLists.txt: 
	* src/core/CMakeLists.txt: 
	* src/core/tnlCudaSupport.h (checkCUDAError):
	  - adding new error codes 
	* src/core/tnlLongVectorCUDA.h: 
	* src/matrix/tnlRgCSRMatrix.h:
	  - fixing copying to device - useAdaptiveGroupSize was not copied 
	* tests/benchmarks/sparse-matrix-benchmark.h: 
	* tests/benchmarks/tnlSpmvBenchmark.h: 
	* tests/benchmarks/tnlSpmvBenchmarkRgCSRMatrix.h: 
	* tnlConfig.h.in:
	  - adding prefix TNL to CONFIG_DIRECTORY 
	* tools/src/tnl-err-norms.cpp: 
	* tools/src/tnl-grid-view.cpp: 
	* tools/src/tnl-matrix-convert.cpp: 
	* tools/src/tnlcurve2gnuplot.cpp: 

73

74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136
2011-06-12  Tomas Oberhuber  <tomas.oberhuber@fjfi.cvut.cz>

    The following chages reflects conversion of the project
    from the autools to CMake. 

	* src/diff/mdiff-config.h: Removed.
	* tests/tnl-sparse-matrix-check.cfg.desc: Removed.
	* tests/tnl-sparse-matrix-test: Removed.
	* tools/share/mgrid-view.cfg.desc: Removed.
	* tools/src/bin2gnuplot.cpp: Removed.
	* tools/src/compare-files.h: Removed.
	* tools/src/debug.h: Removed.
	* tools/src/diff-norm.cpp: Removed.
	* tools/src/functions-benchmark-cuda.cu.h: Removed.
	* tools/src/mdiff-err-norms.cpp: Removed.
	* tools/src/mgrid2gnuplot.cpp: Removed.
	* tools/src/read-file.h: Removed.
	* src/core/low-level/CMakeLists.txt: New file.
	* src/dummy.cpp: New file.
	* tests/sparse-matrix-benchmark.cu: New file.
	* tnlConfig.h.in: New file.
	* tools/share/tnl-grid-view.cfg.desc: New file.
	* buildAll: 
	* CMakeLists.txt: 
	* src/CMakeLists.txt: 
	* src/config/CMakeLists.txt: 
	* src/core/CMakeLists.txt: 
	* src/core/tnlFile.cpp: 
	* src/core/tnlFile.h: 
	* src/core/tnlString.cpp: 
	* src/core/tnlTimerCPU.cpp: 
	* src/core/tnlTimerRT.cpp: 
	* src/debug/CMakeLists.txt: 
	* src/debug/parse.cc: 
	* src/debug/tnlDebugParserbase.h: 
	* src/debug/tnlDebugScanner.cpp: 
	* src/debug/tnlDebugScanner.l: 
	* src/diff/CMakeLists.txt: 
	* src/matrix/CMakeLists.txt: 
	* src/matrix/tnlAdaptiveRgCSRMatrix.h: 
	* src/matrix/tnlFastCSRMatrix.h: 
	* src/matrix/tnlMatrix.h: 
	* src/matrix/tnlRgCSRMatrix.h: 
	* src/mesh/CMakeLists.txt: 
	* src/solver/CMakeLists.txt: 
	* tests/CMakeLists.txt: 
	* tests/sparse-matrix-benchmark.cpp: 
	* tests/sparse-matrix-benchmark.h (main): 
	* tests/tnlSpmvBenchmark.h: 
	* tests/tnlSpmvBenchmarkAdaptiveRgCSRMatrix.h: 
	* tests/tnlSpmvBenchmarkCSRMatrix.h: 
	* tests/tnlSpmvBenchmarkHybridMatrix.h: 
	* tests/tnlSpmvBenchmarkRgCSRMatrix.h: 
	* tests/unit-tests/core/tnlCommunicatorTester.h: 
	* tools/share/CMakeLists.txt: 
	* tools/share/Makefile.am: 
	* tools/src/CMakeLists.txt: 
	* tools/src/tnl-err-norms.cpp (WriteHeader): 
	(main): 
	* tools/src/tnl-grid-view.cpp (main): 
	* tools/src/tnl-matrix-convert.cpp (main): 
	* tools/src/tnlcurve2gnuplot.cpp (main): 

137 138 139 140 141 142 143 144 145
2011-06-09  oberhuber  <oberhuber@yvonne>

	* src/matrix/tnlCSRMatrix.h: 
	* src/matrix/tnlRgCSRMatrix.h: 
	* tests/sparse-matrix-benchmark.h: 
	* tests/tnlSpmvBenchmarkHybridMatrix.h: 
	* tests/tnlSpmvBenchmarkRgCSRMatrix.h:
	 - working on the adaptive groups 

146 147 148 149 150 151 152 153
2011-06-05  Tomas Oberhuber  <tomas.oberhuber@fjfi.cvut.cz>

	* src/matrix/tnlAdaptiveRgCSRMatrix.h: 
	   - better threads to rows distribution for smaller chunks 
	* tests/matrix-formats-test.h: 
	* tests/sparse-matrix-benchmark.h: 
	* tests/tnlSpmvBenchmarkAdaptiveRgCSRMatrix.h: 

154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182
2011-05-29  Tomas Oberhuber  <tomas.oberhuber@fjfi.cvut.cz>

	* tests/cusp-test.cu: Removed.
	* tests/draw-matrices: New file.
	  - creating script for drawing testing matrices
	* src/core/tnlLongVectorCUDA.h:
	  - fixing problem of initiliazing the type Real by 0.0 which does not work for general types 
	* src/core/tnlString.cpp (tnlString::operator +):  
	* src/core/tnlString.h (tnlString): 
	  - adding operator + for const char*
	* src/matrix/tnlAdaptiveRgCSRMatrix.h: 
	* src/matrix/tnlCSRMatrix.h: 
	* src/matrix/tnlMatrix.h (printOut): 
	* src/matrix/tnlRgCSRMatrix.h:
	  - implementing drawing of matrices in Postscript 
	* tests/run-sparse-matrix-benchmark: 
	* tests/sparse-matrix-benchmark.cfg.desc: 
	* tests/sparse-matrix-benchmark.cpp (main): 
	* tests/sparse-matrix-benchmark.h: 
	* tests/tnlSpmvBenchmark.h: 
	* tests/tnlSpmvBenchmarkAdaptiveRgCSRMatrix.h: 
	* tests/tnlSpmvBenchmarkHybridMatrix.h: 
	* tests/tnlSpmvBenchmarkRgCSRMatrix.h: 
	* tools/share/mgrid-view.cfg.desc:
	  - adding new command line arguments for drawing matrices 
	* tools/src/tnl-grid-view.cpp (main): 
	* tools/src/tnl-grid-view.h:
	  - implementing drawing of matrices 

183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199
2011-05-18  Tomas Oberhuber  <tomas.oberhuber@fjfi.cvut.cz>

	* src/matrix/tnlAdaptiveRgCSRMatrix.h:
	  - implementing getElement method  
	* src/matrix/tnlRgCSRMatrix.h:  
	* tests/matrix-formats-test.cpp (main): 
	* tests/matrix-formats-test.h: 
	  - adding test of tnlAdaptiveRgCSRMatrix
	* tests/run-matrix-formats-test: 
	* tests/unit-tests/Makefile.am: 
	* tests/unit-tests/matrix/tnlAdaptiveRgCSRMatrixTester.h (triDiagonalMatrixTest): 
	(upperTriangularMatrixTest): 
	(diagonalMatrixTest): 
	(suite): 
	* tests/unit-tests/tnl-unit-tests.cpp (main):
	  - adding tests of tnlAdaptiveRgCSRMatrix 

200 201 202 203 204 205 206 207 208 209 210 211
2011-05-17  Tomas Oberhuber  <tomas.oberhuber@fjfi.cvut.cz>

	* tests/unit-tests/matrix/tnlAdaptiveRgCSRMatrixTester.h: New file.
	* src/matrix/tnlAdaptiveRgCSRMatrix.h: 
	  - conversion from CSR format
	  - SpMV operation
	* tests/run-sparse-matrix-benchmark: 
	* tests/sparse-matrix-benchmark.h:
	  - changed benchmarks of the RgCSR format with variable group size   
	* tests/unit-tests/tnl-unit-tests.cpp (main):
	  - added test of tnlAdaptiveRgCSRMatrix

212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232
2011-05-15  Tomas Oberhuber  <tomas.oberhuber@fjfi.cvut.cz>

	* tests/cusp-test.h: Removed.
	* tests/tnlSpmvBenchmark.h: New file.
	* tests/tnlSpmvBenchmarkCSRMatrix.h: New file.
	* tests/tnlSpmvBenchmarkHybridMatrix.h: New file.
	* tests/tnlSpmvBenchmarkRgCSRMatrix.h: New file.
	* src/core/tnlLongVectorCUDA.h: Adding operator =, denying copy constructors
	* src/core/tnlLongVectorHost.h: Adding operator =, denying copy constructors 
	* src/matrix/tnlCSRMatrix.h: Adding reorderRows
	* src/matrix/tnlMatrix.h: Adding sortRowsDecreasignly
	* src/matrix/tnlRgCSRMatrix.h:  
	* tests/Makefile.am:  
	* tests/matrix-formats-test.h: 
	* tests/run-sparse-matrix-benchmark: Adding debug
	* tests/sparse-matrix-benchmark.cpp (main): 
	* tests/sparse-matrix-benchmark.h: refactoring
	* tests/unit-tests/Makefile.am:  
	* tests/unit-tests/matrix/tnlRgCSRMatrixTester.h (triDiagonalMatrixTest): 
	(diagonalMatrixTest): 

233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328
2011-03-09  Tomas Oberhuber  <tomas.oberhuber@fjfi.cvut.cz>

	* configure.ac: 
	* src/core/tnlArray.h: 
	* src/core/tnlCommunicator.h: 
	* src/core/tnlCriticalSection.cpp: 
	* src/core/tnlSharedMemory.h: 
	* src/debug/Makefile.am: 
	* src/matrix/tnlRgCSRMatrix.h (bindRgCSRMatrixCUDATexture):
	 - fixing binding vector x to texture
	* tests/convert-matrices: 
	* tests/florida-matrix-market: 
	* tests/Makefile.am: 
	* tests/unit-tests/core/tnlCommunicatorTester.h (testCommunicatorBroadcastLongVector): 
	(testCommunicatorBroadcast): 
	(testCommunicatorScatter): 
	(testCommunicatorGather): 
	(testCommunicatorSendReceive): 
	(testCommunicatorSendReceiveOnLongVector): 
	(testCommunicatorBarrier): 
	(testCommunicatorReduction): 
	* tests/unit-tests/Makefile.am: 

2011-02-27  Tomas Oberhuber  <tomas.oberhuber@fjfi.cvut.cz>

	* src/core/tnlCoordinates.h: New file.
	* src/mesh/tnlDistributedGrid.h: New file.
	* src/core/Makefile.am: 
	* src/core/tnlCommunicator.h (getGoupSize<3>): 
	(getGoupSize<2>): 
	(getGoupSize<1>): 
	* src/core/tnlCriticalSection.cpp (tnlCriticalSection::leave): 
	(tnlCriticalSection::tnlCriticalSection): 
	(tnlCriticalSection::enter): 
	* src/core/tnlCriticalSection.h (tnlCriticalSection): 
	* src/debug/tnlDebugStructure.cpp (tnlDebugStructure::Debug): 
	* src/mesh/Makefile.am: 
	* src/mesh/tnlGrid.h: 
	* tests/unit-tests/core/tnlCommunicatorTester.h (testCommunicatorBroadcastLongVector): 
	(testCommunicatorBroadcast): 
	(testCommunicatorScatter): 
	(testCommunicatorGather): 
	(suite): 
	(testCommunicatorBarrier): 
	(testCommunicatorReduction): 
	* tests/unit-tests/tnl-unit-tests.cpp: 

2011-02-23  Tomas Oberhuber  <tomas.oberhuber@fjfi.cvut.cz>

	* src/core/tnlCommunicator.h: 
	   - implementing method broadcast
	* src/core/tnlSharedMemory.h: 
	   - fixing bug with missing deletion of the shared memory
	* tests/unit-tests/core/tnlCommunicatorTester.h (testCommunicatorSendReceiveOnLongVector): 
	* tests/unit-tests/core/tnlSharedMemoryTester.h (testSharedMemory): 
	* tests/unit-tests/tnl-unit-tests.cpp: 

2011-02-22  Tomas Oberhuber  <tomas.oberhuber@fjfi.cvut.cz>

	* tests/unit-tests/core/tnlSharedMemoryTester.h: New file.
	* src/core/tnlCommunicator.h:
	 - rewriting non-vector send and receive using tnlSharedMemory 
	* src/core/tnlSharedMemory.h: 
	 - finishing implementation
	* tests/unit-tests/core/tnlCommunicatorTester.h (testCommunicatorSendReceiveOnLongVector): 
	* tests/unit-tests/tnl-unit-tests.cpp (main): 

2011-02-22  oberhuber  <oberhuber@yvonne>

	* src/core/tnlSharedMemory.h: New file.
	 - starting implementation of the shared memory 
	* src/core/tnlCommunicator.h: 

2011-02-20  Tomas Oberhuber  <tomas.oberhuber@fjfi.cvut.cz>

	* src/core/tnlCriticalSection.cpp: New file.
	* src/core/tnlCriticalSection.h: New file.
	  - implementing tnlCriticalSection
	* src/core/Makefile.am: 
	* src/core/tnlCommunicator.h: 
	  - rewriting the old code using tnlCriticalSection
	  - implementing send and receive for vectors
	* src/debug/tnlDebug.h: 
	* tests/unit-tests/core/tnlCommunicatorTester.h (testCommunicatorSendReceiveOnLongVector): 
	(suite): 

2011-02-19  Tomas Oberhuber  <tomas.oberhuber@fjfi.cvut.cz>

	* src/core/tnlCommunicator.h: 
	 - implementing methods start, send and receive
	* tests/unit-tests/core/tnlCommunicatorTester.h (testCommunicatorInitiation): 
	(testCommunicatorSendReceive): 
	(suite): 

2011-02-09  oberhuber  <tomas.oberhuber@fjfi.cvut.cz>

329
	* configure.ac: 
330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395
	* src/config/Makefile.in: 
	* src/core/low-level/cuda-long-vector-kernels.h: 
	 - changing number of threads per block for the CUDA reduction to get occupancy 1.0
	* src/matrix/tnlCSRMatrix.h: 
	* src/solver/tnlMersonSolver.h: 
	 - optimizing the updating kernel to get occupancy 1.0 

2011-02-06  Tomas Oberhuber  <tomas.oberhuber@fjfi.cvut.cz>

	* src/core/debug.h: Removed.
	* src/diff/debug.h: Removed.
	* src/diff/mdiff-debug.h: Removed.
	* tools/src/debug.h: Removed.
	* src/config/Makefile.am: 
	   - removing debug.h 
	* src/config/tnlConfigDescriptionParser.cpp: 
	* src/core/Makefile.am:
	   - removing debug.h 
	* src/core/tnlObject.cpp:
	   - replacing debug.h with tnlDebug.h 
	* src/core/tnlString.cpp:
	   - replacing debug.h with tnlDebug.h 
	* src/debug/Makefile.am:
	   - removing debug.h 
	* src/debug/tnlDebug.cpp (tnlInitDebug):
	   - the flag debug is set to yes even if the parsing fails 
	* src/debug/tnlDebug.h: 
	   - adding prefix #TNLDBG# to all printouts
	* src/diff/Makefile.am:
	   - removing debug.h 
	* tools/src/Makefile.am:
	  - removing debug.h 

2011-02-05  Tomas Oberhuber  <tomas.oberhuber@fjfi.cvut.cz>

    * src/core/tnlComminucator.h New file.
       - starting implementation of the CUDA/OpenCL/MPI communication framework
	* src/matrix/tnlCSRMatrix.h: 
	   - implementing method performSORIteration
	* src/solver/tnlMatrixSolver.h: 
	   - adding method getResidue (moved from the tnlSORSolver.h)
	* src/solver/tnlSORSolver.h:
	   - refactoring the solver to work with the new system o templates  



December 2010 oberhuber  <tomas.oberhuber@fjfi.cvut.cz>
* Explicit solvers now have device where the data are stored as a parameter too.
* tnlGrid  is now a template with dimensions as parameter.
  This avoids classes like tnlGrid1D, tnlGrid2D etc.
  The device where the grid lives is also template parameter.
* Implemented some basic operations lika Saxpy norms evaluation etc. on tnlLongVector.
* Implemented tnlLongVector having the device where it lives as parameter.
  At this moment it only supports CUDA. This should unified code for CPU and GPU
  as much as possible.

November 2010 oberhuber  <tomas.oberhuber@fjfi.cvut.cz>
* Implemented tnlFile. It supports bzip2 compression transparently.
  It makes all I/O operation simpler. It also allows to save data directly
  from GPU devices - also for simplifing use of large numerical structures.

July 2010 oberhuber  <tomas.oberhuber@fjfi.cvut.cz>
* Implementing sparse matrices in CUDA - Coalesced CSR format. 

December 2009 oberhuber  <tomas.oberhuber@fjfi.cvut.cz>
* Starting implementation of the CUDA support.