aboutsummaryrefslogtreecommitdiff
path: root/Carpet
Commit message (Collapse)AuthorAge
...
* CarpetLib: Support padding in enovol prolongation operatorErik Schnetter2012-11-22
|
* Carpet: Turn time level / prolongation operator warning into an info messageErik Schnetter2012-11-22
| | | | This warning has many false positives, and is informational only.
* Carpet: Simplify codeErik Schnetter2012-11-22
|
* CarpetLib: Make th::time_interpolation_during_regridding publicErik Schnetter2012-11-22
|
* CarpetLib: Set up all timers on all processes all the timeErik Schnetter2012-11-22
| | | | Set up timers even on processes that won't use them, so that reducing timers across processes works.
* CarpetLib: Re-enable OpenMP collapse clausesErik Schnetter2012-11-22
|
* Allow padding in transport operatorsErik Schnetter2012-11-22
| | | | | Rewrite padding infrastructure. Add padded array extents to transport operator APIs.
* LoopControl: Add missing copyprivateErik Schnetter2012-11-22
|
* CarpetLib: add volume averaged cc eno operatorRoland Haas2012-11-20
|
* CarpetLib: Define dh members tallying the size of refinement boundariesErik Schnetter2012-11-14
| | | | Define dh members tallying the size of refinement boundaries. This may be used to implement (2d) face grid functions at some point.
* CarpetMask: Add missing OpenMP reduction clauseErik Schnetter2012-11-06
|
* CarpetReduce: Do not access NULL pointerErik Schnetter2012-11-06
|
* CarpetReduce: Use self-defined MPI operators for complex numbersErik Schnetter2012-11-06
| | | | Do not treat complex numbers as two real numbers.
* CarpetReduce: Avoid duplicate local variablesErik Schnetter2012-11-06
|
* CarpetLib: Add self-checks to MPI reduction operatorsErik Schnetter2012-11-06
|
* CarpetLib: Use correct MPI reduction operator for product reductionsErik Schnetter2012-11-06
|
* Carpet: Correct switch statement in CarpetMPIDatatypeErik Schnetter2012-11-06
|
* CarpetRegrid2: Use "isfinite" instead of "not isnan"Erik Schnetter2012-11-05
|
* Carpet: Correct commentErik Schnetter2012-11-01
|
* CarpetIOHDF5: fix check for missing datasets when called in FileReader modeRoland Haas2012-10-29
| | | | only check datasets whose variables were actually requested to be read
* Carpet: Set NPROCS for kasner test casesErik Schnetter2012-10-28
|
* Add a timer tree for measuring time spent in each Carpet modeIan Hinder2012-10-26
| | | | | | * This can be used for determining which mode, or which refinement level or map, is taking up a lot of time in a simulation * Add parameters to control whether local mode and each map are timed * Time the transitions between modes
* Support more than one hierarchy of timersIan Hinder2012-10-26
| | | | | | * Introduce a TimerTree class which contains a pointer to the root and current TimerNodes of that tree * Adapt TimerNode and Timer classes to use TimerTrees rather than static variables * Declare a main_timer_tree variable, which is the default tree to use if you don't specify one in Timer timer("timername")
* Carpet: Add test cases checking cctk_time during initial data setupErik Schnetter2012-10-25
| | | | Add analytic test cases checking cctk_time while setting up initial data on past timelevels.
* CarpetIOHDF5: close all HDF5 objects when output file is closedRoland Haas2012-10-24
| | | | also garbage collect HDF5 at each H5close
* CarpetLib: Update cycle.h from upstreamErik Schnetter2012-10-24
|
* Carpet, CarpetLib: Move #define _GNU_SOURCE to correct fileErik Schnetter2012-10-24
|
* Carpet: Execute CPINITIAL only after initial data setup, not after recoveringErik Schnetter2012-10-24
|
* Carpet, CarpetRegrid2: When comparing refined regions treat non-existing ↵Erik Schnetter2012-10-24
| | | | levels as empty
* CarpetRegrid2: Correct snapping to coarse grid for cell centeringErik Schnetter2012-10-23
| | | | Ensure that small, thin bboxes don't vanish.
* CarpetRegrid2: Perform domain decomposition only when a refinement level changesErik Schnetter2012-10-23
| | | | Compare new to old refined regions for each refinement level. Don't perform domain decomposition if a level didn't change. If no level changed, tell Carpet.
* Carpet: Requirements: Add precondition to clause_t::active_on_timelevelErik Schnetter2012-10-23
|
* Carpet: Requirements: Warn about unnecessary syncsErik Schnetter2012-10-23
|
* Carpet: Correct errors in adaptive step size handlingErik Schnetter2012-10-23
|
* CarpetLib: Remove restrict qualifier from pointer castErik Schnetter2012-10-23
|
* Carpet: Re-indent timer macrosErik Schnetter2012-10-23
|
* Carpet: Requirements: Add aliased function to invalidate variablesErik Schnetter2012-10-23
|
* Carpet: Correct #ifdefs in CarpetSimpleMPIDatatypeLengthErik Schnetter2012-10-22
|
* Remove Carpet's mechanisms to deal with how to call isnanErik Schnetter2012-10-22
| | | | Remove Carpet's mechanisms to deal with how to call isnan. Instead, expect isnan in std::, and rely on Cactus to correct things if this is not the case.
* Merge branch 'master' of carpetcode.org:carpetChristian Reisswig2012-10-20
|\
| * CarpetIOHDF5: output index files for sliced output if requestedRoland Haas2012-10-20
| |
| * CarpetIOASCII: do not output coordinate column labels for grid arrays inRoland Haas2012-10-20
| | | | | | | | | | | | | | | | compact_format since the coordinates themselves are not output either. Actually *only* output coordinate column labels for grid functions.
| * LoopControl: Change printstats default to "no"Erik Schnetter2012-10-19
| |
* | CarpetInterp: New ENO2 interpolator.Christian Reisswig2012-10-20
|/
* Carpet: Correct errors with adaptive time steppingErik Schnetter2012-09-27
|
* Introduce parameter "additional_overlap_zones" to control the amount of ↵Christian Reisswig2012-09-26
| | | | | | | | additional (user-requested) overlap zones. This is similar to additional_buffer_zones. Also fixed a bug in reporting the number of overlap zones in stdout. ("buffers" instead of "overlaps") was used.
* CarpetLib: Clean up backtrace generationErik Schnetter2012-09-18
| | | | Also add test parameter file.
* Carpet: Remove outdated backtrace function declarationsErik Schnetter2012-09-18
|
* Carpet: fix mg level count typo during intializationRoland Haas2012-09-18
| | | | | this caused the initial time levels to always have t=0 associated with them
* CarpetIOASCII: Implement using grid::coordinates to output coordinatesErik Schnetter2012-09-18
|