aboutsummaryrefslogtreecommitdiff
path: root/Carpet
Commit message (Collapse)AuthorAge
* Carpet: Pass MPI communicators as pointers instead of intsErik Schnetter2005-08-26
| | | | | | | | | Pass MPI communicators as pointers instead of ints because they may not be implemented as ints. Use MPI_COMM_NULL instead of -1 as illegal communicator. darcs-hash:20050826144511-891bb-324a5d22efcbbd53cac35899e30559b06cfdfdb6.gz
* CarpetInterp: optimise global interpolator communication schemeThomas Radke2005-08-26
| | | | | | | | | | | | CarpetInterp used to use CarpetLib's data class to exchange interpolation information (interpolation coordinates and a source map as inputs, interpolation results and status/return codes as outputs) between processors. This point-to-point communication has been replaced by explicit collective MPI operations which should now make global interpolations faster. For a general overview on the implementation see CarpetInterp's thorn documentation. darcs-hash:20050826115157-776a0-d910b51d7a26cef12e13408a79f11ed2826f5ed1.gz
* Carpet: Allow multi-model simulationsErik Schnetter2005-08-25
| | | | | | | | | | Split MPI_COMM_WORLD according to a new parameter Carpet::model into different sets, and evolve these sets independently. Add aliased functions to query the MPI communicators for this simulation and for all models. darcs-hash:20050825084739-891bb-7f102b46ec495f884be5197eaf077eef93f89dac.gz
* CarpetLib: Initialise Carpet's MPI communicator with an arbitrary communicatorErik Schnetter2005-08-25
| | | | | | | Initialise Carpet's MPI communicator not with MPI_COMM_WORLD, but with a communicator that is passed in. darcs-hash:20050825084335-891bb-38744ff9cbfb5349e34676897dba3356ef7e462e.gz
* CarpetWeb: Update link to Whisky web pagesErik Schnetter2005-08-24
| | | | darcs-hash:20050824145010-891bb-0173d254969a5c9123aae3882f461d03434f936c.gz
* CarpetWeb: Add August 2005 status reportErik Schnetter2005-08-24
| | | | darcs-hash:20050824144936-891bb-4519b3de89acb2fa1a0fe82443fab9cece3d0549.gz
* CarpetInterp: Remove commented out assertErik Schnetter2005-08-24
| | | | darcs-hash:20050824113516-891bb-b3b73b0c0bc89ad54bd64d6919621e7a76d6accc.gz
* CarpetIOHDF5: bugfix for outputting the same variable into multiple filesThomas Radke2005-08-23
| | | | | | | | | | Before a variable is output it is checked whether it has been output already during the current iteration (eg. due to triggers). This check was only variable-based and therefore caused problems when the same variable was to be output to multiple files (using different alias names). Now the check has been extended to also take the output filenames into account. darcs-hash:20050823135345-776a0-1555987b4aee34bb646e67f491375dbcc44dddad.gz
* CarpetLib: bugfix in collective commbuffer schemeThomas Radke2005-08-16
| | | | | | | | | The low-level routine gdata::change_processor() wasn't aware of the collective commbuffers communication scheme. This caused problems with CarpetInterp which still uses this function in order to communicate data. But not for long anymore, another patch is waiting already to be pushed. darcs-hash:20050816135442-776a0-60371ebba14505a9d402bb08a9b0696d4b3d4702.gz
* Carpet*: generalise the comm_state class for collective buffer communicationsThomas Radke2005-08-15
| | | | | | | | | CarpetLib's comm_state class (actually, it's still just a struct) has been extended to handle collective buffer communications for all possible C datatypes at the same time. This makes it unnecessary for the higher-level communication routines to loop over each individual datatype separately. darcs-hash:20050815150023-776a0-dddc1aca7ccaebae872f9f451b2c3595cd951fed.gz
* CarpetEvolutionMask: Handle outer buffer zonesErik Schnetter2005-08-12
| | | | | | Do not shrink the domain if the buffer zones are outer. darcs-hash:20050812143317-891bb-5e95257cd7a5685fbcc02985e5802e3f1e070a78.gz
* CarpetLib: Implement fast recomposingErik Schnetter2005-08-11
| | | | | | | | For each refinement level that is to be recomposed, check whether it has the same structure as before, and if so, do nothing. This is controlled by a new flag CarpetLib::fast_recomposing. darcs-hash:20050811120347-891bb-f937c21ddeac7d909cae41d487e9fd74a5ce8cc8.gz
* CarpetInterp: Remove support for optional argument "interpolation_times"Erik Schnetter2005-08-11
| | | | | | | | | | The optional argument "interpolation_times" allowed interpolating not at the current time (cctk_time), but at a slightly earlier time, if there were enough time levels present. This feature was unused and broken. darcs-hash:20050811170146-891bb-db58ea38ca8160729f9c8a53d17db2381dc27535.gz
* CarpetInterp: Test also time derivatives in the test casesErik Schnetter2005-08-11
| | | | darcs-hash:20050811163508-891bb-a25e3c59937a75684218fde1cf4fafd19087f2c9.gz
* CarpetInterp: Update waveinterp test casesErik Schnetter2005-08-11
| | | | | | | Update the waveinterp test cases after correcting three timelevel initialisation. darcs-hash:20050811113402-891bb-b40f5401b44d626bda9611de2622a77395c16295.gz
* Carpet: Clean up outer-buffers.par fileErik Schnetter2005-08-10
| | | | darcs-hash:20050810192718-891bb-91b71ce58755065f6846a5bd40686c9953f3d1a6.gz
* Carpet: Correct three timelevel initialisationErik Schnetter2005-08-10
| | | | | | | | There was another error in the three timelevel initialisation. The first backwards time step did not have enought coarser level time steps for the necessary prolongation operations. darcs-hash:20050810192602-891bb-ba9ed7d58dea6a3ddbf59be0fa6baa1ede509963.gz
* CarpetInterp: Use explicit source map specifications even in single map modeErik Schnetter2005-08-10
| | | | darcs-hash:20050810170009-891bb-168facf88ab10cc5c1732d037527a66bc134453c.gz
* CarpetInterp: Make time_deriv_order per output array instead of per grid pointErik Schnetter2005-08-10
| | | | | | | Interpret the option table parameter time_deriv_order per output array instead of per grid point. darcs-hash:20050810165758-891bb-a4b84c90771da42c20878ebd59a797ef44a29b78.gz
* CarpetInterp: Remove commented out codeErik Schnetter2005-08-10
| | | | darcs-hash:20050810170107-891bb-bc5b8fa8a90d7289c0c7d76f3ad4a00aa3c1038f.gz
* CarpetEvolutionMask: New thornErik Schnetter2005-08-10
| | | | | | | | Provide a grid function "evolution_mask" which is zero for all grid points that do not need to be evolved because they are going to be restrict from a finer grid. darcs-hash:20050810125908-891bb-3a6441b6cd41b7bafda6a42c672925ac98fbb668.gz
* CarpetLib: typo in push of << operatorSteve White2005-08-09
| | | | | | | | | This fixes a typo in the previous push. I also took the opportunity to bring the patch into better compliance with the coding style. darcs-hash:20050809143402-90671-5feae89e34e29f1648b31a66d97d578a5b283f1b.gz
* CarpetInterp: Correct time interpolation stencilsErik Schnetter2005-08-11
| | | | darcs-hash:20050811170351-891bb-b7c586cb8494f23fac7254f169a1a2aeec523f97.gz
* CarpetLib: give data an << operatorSteve White2005-08-09
| | | | | | | | | | Added a ostream & << operator for the data class. Made an explicit instantiation for CCTK_REAL. Usually the amount of data renders such an operator useless, but in test programs it is handy to be able to just print the data values. darcs-hash:20050809093318-90671-dd30c55dc8757efa96a9bfbe90a48ff445e26f07.gz
* Carpet: Add test case for outer buffer zonesErik Schnetter2005-08-09
| | | | darcs-hash:20050809093357-891bb-88b8f8019ddd2c088e8fbe326f45a9d0eea34114.gz
* Carpet: Use outer buffer zonesErik Schnetter2005-08-09
| | | | | | | | | | | | | | | | Deprecated the previous buffer zone mechanism. Introduce new parameters Carpet::use_outer_buffer_zones and Carpet::num_integrator_substeps. The number of buffer zones are then calculated automatically, and are added to the outer edge of the grid instead of reducing the domain. Adapt to the API change of the dh class. Both new and old style buffer zones can exist at the same time, although I would not do that. darcs-hash:20050809091707-891bb-c25d68de3b04b03c06078aaee39f389b67f1416a.gz
* CarpetLib: Add outer buffer zonesErik Schnetter2005-08-08
| | | | | | | | | | | | | | | | | | | | | Rename the previous buffer zones to "inner buffer zones". Introduce additional "outer buffer zones". Their meaning is the same, except that inner buffer zones are taken from the computational domain, making it smaller, while outer buffer zones are added to the outside, similar to ghost zones. This makes them easier to handle, both internally in Carpet and for the end user. This changes the API of the dh class. There is a new field "is_interproc" in the "dboxes" structure; it specifies whether the whole boundary is an interprocessor boundary. (Note that boundaries can be partly interprocessor and partly refinement boundaries.) Both kinds of buffer zones can exist at the same time, although I would not do that. darcs-hash:20050808200647-891bb-9d9a8eefaf7bcb665d09869a8b564f3769d2ecc2.gz
* Carpet: Correct problems with negative time stepsErik Schnetter2005-08-09
| | | | | | | | | | | | When converting from physicsl time steps (delta_time) to Carpet's time steps (th::delta_time), do not divide by abs(delta_time), but by delta_time instead. When evolving backwards in time while initialising three time levels, change the sign of the physical time steps only, not the sign of Carpet's time steps. darcs-hash:20050809084940-891bb-2517b1568696c71278a98db6261515817a90247a.gz
* Carpet: Handle negative time steps correctly during prolongationErik Schnetter2005-08-08
| | | | darcs-hash:20050808194745-891bb-74c3f74d1b5a5886ea63e11671aa7f425d5a8daa.gz
* CarpetIOASCII: Repair outputting all time levelsErik Schnetter2005-08-08
| | | | | | Use positive time levels when outputtting all time levels. darcs-hash:20050808180907-891bb-7fdf3f70084dbcd9bc49d6b50869de8774032805.gz
* CarpetInterp: fix waveinterp testsuiteThomas Radke2005-08-08
| | | | | | | | | | | | | The waveinterp testsuite was broken, many output files had different values. Also, the old output files didn't contain the recently introduced header comment lines saying which thorn did output them. The waveinterp testsuite has now been split into two separate tests, waveinterp-1p and waveinterp-2p to run on 1 and 2 processors respectively. The only differences are in the CarpetIOASCII files which contain either 1 or 2 components in the output. darcs-hash:20050808164941-776a0-fbecb78b32c48a16f030996afcf1259276c0a0a0.gz
* CarpetIOASCII: Output DISTRIB=const variables only on one processorErik Schnetter2005-08-08
| | | | | | | | | DISTRIB=const variables are supposed to be identical on all processors. Output them only on the root processors to save space. This makes it also possible to compare output from different numbers of processors in test cases. darcs-hash:20050808114339-891bb-f058e467a9b95610675245bdb6dfa2cfa3104c6c.gz
* CarpetLib: Change default settings for the communication tuning parametersErik Schnetter2005-08-08
| | | | | | | These default settings are now believed to be sensible and safe for everybody. darcs-hash:20050808132929-891bb-48231878e0a5ea02312823f4b96cad1c79fdba9f.gz
* CarpetLib: Correct typo in commentErik Schnetter2005-08-08
| | | | darcs-hash:20050808132501-891bb-e03ca26e0adfe32f01065ecae0b987785855190a.gz
* CarpetLib: Change indentation of typestring definitionsErik Schnetter2005-08-08
| | | | darcs-hash:20050808132424-891bb-e352aafd9745fb66df45212a4a28e3175a7b58f0.gz
* CarpetLib: Introduce new types b2vect and i2vectErik Schnetter2005-08-08
| | | | | | | | | | | | These types are typedef vect<vect<bool,dim>,2> b2vect; typedef vect<vect<int,dim>,2> i2vect; They are similar to bbvect and iivect, but have the order reversed. They are useful if you need one element per face of a grid variable. darcs-hash:20050808132221-891bb-fa40b6d4fb8e760c7005adf7e526dd0a6597f0ec.gz
* CarpetLib: Disable check for time extrapolation for op_copy grid functionsErik Schnetter2005-07-26
| | | | | | | | | | Disable the check for time extrapolation for grid functions that are not interpolated but copied (transport_operator=op_copy). This check needs to know the sign of delta_time, and that sign is not known here. (For example, 3-timelevel-initialisation evolves both forwards and backwards in time.) darcs-hash:20050726111122-891bb-0c2130f2dfdd86c25575cbb88cb7c8ad80089eac.gz
* CarpetReduce: Change to new style C++ header filesErik Schnetter2005-07-26
| | | | darcs-hash:20050726110949-891bb-fe7d442ee41c585b7078fd6746840cd68de1d770.gz
* Carpet: Make 3-timelevel-initialisation work againErik Schnetter2005-07-25
| | | | | | | | | | | 3-timelevel-initialisation was broken: the refinement levels were initialised in the wrong order, so that the method was formally only first order convergent. Reorder how the individual steps of the algorithm traverse the refinement levels. darcs-hash:20050725212427-891bb-72d0e8554e98e7126187b61ef87fe447085dc4af.gz
* Carpet: Add more const qualifiers in Initialise.ccErik Schnetter2005-07-25
| | | | darcs-hash:20050725122347-891bb-bb71264acc713fce942f5409592f543f2eece03b.gz
* Carpet: Do not pass Cactus parameters aroundErik Schnetter2005-07-25
| | | | darcs-hash:20050725121823-891bb-0e8ca67a430b353f20100f562b1455b309393511.gz
* Carpet: Add some const qualifiers to function argumentsErik Schnetter2005-07-25
| | | | darcs-hash:20050725121646-891bb-829f88b403542b5be5bcfc3d1aa41b2d57c46ade.gz
* Carpet: Make refinement level loops in Initialise.cc more explicitErik Schnetter2005-07-25
| | | | | | | Move some loops out of functions and into the caller. This makes it easier to see which levels are walked in what order. darcs-hash:20050725121402-891bb-0997ea576c96edba7ee53cc7df85a5994c3f6d15.gz
* Carpet: Restructure the file Initialise.ccErik Schnetter2005-07-24
| | | | | | | Give some local functions better names and reorder them. No change in functionality. darcs-hash:20050724114350-891bb-7610f8d3b11165ad04fc65e3e8c6f7d869b8d474.gz
* Carpet: Correct indentation in Initialise.ccErik Schnetter2005-07-23
| | | | darcs-hash:20050723214656-891bb-51db2a158414910d2ee942ce8bda00c9f6588e9a.gz
* CarpetInterp: Rename return valueErik Schnetter2005-08-01
| | | | | | | Do not resize source_map twice. Rename Util_TableGetIntArray return value from ierr to iret. darcs-hash:20050801124730-891bb-4b25ee01ef6a3a21d75a19bc1500eec4ccaf310f.gz
* CarpetIOASCII: Prepend column numbers to column descriptionsErik Schnetter2005-07-30
| | | | | | | Prepend column numbers to column descriptions in the header of the files. darcs-hash:20050730073824-891bb-54139304b2c3da5a767f6f1a6a357234bb6f6c74.gz
* CarpetIOScalar: Prepend column numbers to column descriptionsErik Schnetter2005-07-30
| | | | | | | Prepend column numbers to column descriptions in the header of the files. darcs-hash:20050730073713-891bb-b3588e78345c9ffacff6c314cd4f6869e4a42e04.gz
* Carpet: update parfiles in the par/ directoryThomas Radke2005-07-28
| | | | | | | | | | | | The parfiles didn't run because not all thorns were activated. test_rad.par activated IOASCII and set an IOASCII parameter instead of using CarpetIOASCII. multipatch.par was a very old parfile which didn't run at all. It was therefore removed. This closes bug report http://bugs.carpetcode.org/show_bug.cgi?id=78 "Fails to run test par files (unactivated thorns)". darcs-hash:20050728171716-776a0-104c9ba1f0ff601febbcc682ff60c43edcaba5f0.gz
* CarpetIOHDF5: optimise I/O accesses for the case when all processors recover ↵Thomas Radke2005-07-28
| | | | | | | | | | | | | | from a single chunked checkpoint file The list of HDF5 datasets to process is now reordered so that processor-local components are processed first. When processing the list of datasets, only those will be reopened from which more data is to be read. The check when a variable of a given timelevel has been fully recovered was improved. This closes http://bugs.carpetcode.org/show_bug.cgi?id=87 "Single-file, many-cpu recovery very slow". darcs-hash:20050728122546-776a0-21dfceef87e12e72b8a0ccb0911c76066521e192.gz
generated by cgit v1.2.3 (git 2.39.1) at 2024-07-06 22:46:51 +0000