This is an automated email from the git hooks/post-receive script. It was generated because a ref change was pushed to the repository containing the project "McLachlan".
The annotated tag, ET_2013_05_v0 has been created at 44b99a035ddc3d6168757d381d8a99ee8eceb52b (tag) tagging 966f1fdcffb6c327d4083a285edf895e3cd51da3 (commit) tagged by Frank Löffler on Mon May 27 17:22:44 2013 -0500
- Log ----------------------------------------------------------------- ET_2013_05_v0 tag
Anton Khirnov (1): McLachlan_BSSN: fix the Hamiltonian constraint for CCZ4.
Barry Wardell (27): Enable Vectorisation. Regenerate code with Vectorisation enabled. Enable Vectorisation. Merge branch 'master' into CCZ4 Merge branch 'master' into CCZ4 Regenerate thorns. Add formulation parameter to select between BSSN and CCZ4. Add some regression tests for CCZ4, modelled on the BSSN tests. Make sure boundary conditions and tags are set for CCZ4 Theta. Add some sample CCZ4 parameter files. Generate a separate thorn for CCZ4. Remove CCZ4 specific code from BSSN helper thorn. Add CCZ4 Helper thorn. Rename CMBSSNW and CMBSSNphi to CMW and CMphi. Regenerate thorns. Move CCZ4 regression tests and update parameter files to use the ML_CCZ4 thorn. Merge branch 'CCZ4' Add CCZ4 thorns to 'clean' make target. Add all WaveToy thorns to 'clean' make target. Add testsuites using EinsteinExact. Regenerate code (kranc commit f4d2482bef5b7daf9fc89279db2ddc7ec8ea5bb2) Regenerate code after fix to CCZ4 constraint calculation. Include Hamiltonian and momentum constraints in CCZ4 tests. Regenerate thorns which were missed in 45bcf7046474c39dc6fb7c3e7729cdd1143a70fe. Regenerate CCZ4 Hamiltonian and momentum test results with corrected version of code. Increase tolerance in ML_CCZ4 test cases. Remove tests for _MP and _O8 thorns which no longer exist.
Bernard Kelly (1): McLachlan_BSSN.m: Correct harmonic shift for conformalMethod = phi
Christian D. Ott (2): A sample par file for evolving a single Kerr-Schild BH with 7 levels. * white space change to test commit messages
Daniela Alic (1): Implement CCZ4 formulation.
Erik Schnetter (230): Import initial version Add ML_ prefix. Add urls for Cactus and Kranc. The code also requires Carpet. Add script to re-generate thorns. Add sample parameter file Restructure arrangement Add BSSN RHS. Remove old files Add gauge conditions Add Kerr-Schild parameter files Internal changes, add debug statements Remove debugging statements Add .gitignore file Update .gitignore file Update parameter files Continued development on the BSSN constraints Correct errors in equations Implement arbitrary finite differencing orders Abort code generation script on errors Merge branch 'master' of ssh://relativity.phys.lsu.edu/home/perturbed/gitroot/McLachlan Add comment to BSSN constraint calculations Update autogenerated files Begin to add Tmunu terms Ignore CVS backup files Add local tensor objects for T_ab Remove underscores from HydroBase variable names Use updated Kranc version Add Kranc generated WaveToy Merge branch 'master' of ssh://relativity.phys.lsu.edu/home/perturbed/gitroot/McLachlan Add first-order WaveToy. Use makefile to run Kranc Ignore all temporary thorns Rename MLFO_WaveToy to ML_FOWaveToy Inherit WaveToy from MultiPatch Correct initialisation of dtlapse Add example parameter files for gauge wave Make it more convenient to use global derivatives Correct description and default value for parameter harmonicN Activate TmunuBase Make single black hole evolution work Regenerate WaveToy thorns Use only 2 time levels for evolved variables Merge branch 'master' of ssh://relativity.phys.lsu.edu/home/perturbed/gitroot/McLachlan Regenerate code after removing leftover debugging code from Kranc Add new thorn ML_BSSN_Helper Improve ML_BSSN schedule Update generated code Use 3 time levels by default. Use ADMBase initial data by default Update McLachlan. ML_BSSN_Helper is now required at run time Update Rename trA to trAt Don't calculate ADMBase variables on the boundaries Take note of arXiv:gr-qc/0104063 in the README file Add parameter file for QC-0 evolution Update path to Kranc Re-generate all code Re-generate interface.ccl after correcting error in Kranc Don't schedule boundary conditions explicitly after setting up initial data, Re-generate code after latest change Synchronise ADMBase variables after converting from BSSN to ADM Create several specialised thorns: ML_ADM and ML_BSSN as standard cases, Update ML_BSSN parameetr files. Update path to Kranc installation Update generated ADM thorn Update script to generate BSSN code: Update auto-generated code Re-introduce the distinction between vacuum and matter spacetimes. Add new parameter calculate_ADMBase_variables_at, which determines whether Set CartGrid3D symmetries of McLachlan's variables Correct error in schedule group name Do not register the grid function symmetries in the helper thorns since Update comment Regenerate the code Convert to ADM variables on all grid points, including the boundary. Reformat README Add experiment towards a hydro code Add semicolons to prevent output Add commented out functions for modifying calculations. Add semicolons to prevent output Merge branch 'master' of ssh://relativity.phys.lsu.edu/home/perturbed/gitroot/McLachlan Resolve merge commits Update generated WaveToy Updated generated hydro code Add Helper thorns that went missing during a merge (my apologies) Create helper thorns automatically from a prototype Use correct one-sided stencils at boundaries. Regenerate code Re-generate code after correcting a code generation error in Kranc Register ADM variables as constrained variables, so that MoL copies them Add some documetation, explaining how to check out McLachlan and related Cactus thorns, how to build, and how to run a simple parameter file with McLachlan. Update Add parameter MinimumLapse to enforce a positive lapse value. Correct parameter file and thorn list. Register constrained (e.g. ADMBase) variables with MoL Add missing autogenerated file Add new RHS boundary condition NewRad, implemented in the thorn LSUThorns/NewRad. Rename FOWaveToy to WaveToyFO Add radiative boundary conditions Update documentation Introduce new RHS boundary condition "static" that sets the RHS boundary explicitly to zero. Update auto-generated code Update example parameter files Allow 4 RHS time levels. Add description for converting BSSN_MoL parameter files to McLachlan Do not build the BSSN_W formulation any more, since it is now a part of the standard BSSN formulation. Add new run-time parameter "useMatter", so that the matter terms can be enabled or disabled at run time, instead of requiring a formulation of their own Update auto-generated code Update example parameter files: Remove superfluous synchronisation of RHS variables. Regenerate code Update example parameter file Add CCATIE example for comparison (does not work yet) Commit changes that should have been committed earlier Avoid division-by-zero problems when ShiftGammaCoeff=0 Remove commented out code Merge branch 'master' of /Users/eschnett/Cbeta/arrangements/McLachlan Create CVS password file for anonymous checkout Add McLachlan's dtlapse and dtshift to output variables Handle the case correctly where ShiftGammaCoeff=0. Don't dividy by ShiftGammaCoeff in this case; instead, initialise the (unused) B^i to zero. Use relative instead of absolute path names pointing to Kranc Split RHS evaluation into two routines to reduce instruction cache usage. Correct synchronisation problems Do not generate a separate "matter" thorn; instead, the standard thorn does now handle matter terms Distinguish between persistent variables and those that need to be checkpointed; e.g. eta (the shift driver parameter) is persistent, but is not checkpointed. Add thorn that evaluates the ADM constraints Add thorn that evaluates the kernels for global ADM quantites (ADM mass and ADM angulare momentum) Update auto-generated code Merge /Users/eschnett/Cbeta/arrangements/McLachlan Merge Treate Tmunu variables as tensors, instead of accessing them via local helper scalars. Remove useMatter parameters which are now unused. Regenerate code after correcting error in Kranc. Use new schedule group MoL_PseudoEvolution for the ADM constraints and the ADM quantities. Add second order accurate versions of all thorns Make eta (the Gamma driver damping parameter) a local variable only, not a grid function Generate and add 8th order version as well Schedule boundary conditions in ML_PseudoEvolution Correct boundary conditions Use Ian's version of Kranc, which is found at Don't build or schedule old RHS routine any more Add SpatialShiftGammaCoeffRadius to reduce ShiftGammaCoeff outside of a certain radius Merge branch 'master' of /Users/eschnett/Cbeta/arrangements/McLachlan Don't use explicit path name when including Slicing.h Add dissipation terms to the RHS. Regenerate code Update example parameter files Add benchmark parameter file Add missing definition of temporary variable Regenerate code Regenerate code Correct error in Makefile. Merge branches 'master' and 'master' of carpetcode.dyndns.org:McLachlan Merge branches 'master', 'master', 'master', 'master', 'master', 'master', 'master', 'master', 'master', 'master', 'master', 'master', 'master', 'master' and 'master' of carpetcode.dyndns.org:McLachlan Update path to Kranc Change AssertSymmetricDecreasing to AssertSymmetricIncreasing Use boundary_width instead of 1 when selecting boundary conditions Schedule ADMBase variable calculations in MoL_PseudoEvolution Regenerate code Regenerate code after setting the default tensor weight in Kranc to 0 par: Correct output variables in example parameter files ML_BSSN: Simplify handling of upwinding derivatives Do not apply constraint boundaries manually; Kranc does this now automatically. Do not fix up the boundaries of analysis quantities any more; Kranc now handles this correctly. Do not schedule a group for the constraint boundary conditions any more Do not schedule the RHS calculation at postinitial any more Correct boundary conditions for ADMConstraints and ADMQuantities Regenerate code Do not schedule the conversion to ADMBase at post_recover_variables if they are scheduled in MoL_PostStep, since MoL_PostStep is already scheduled in post_recover_variables. Calculate the Ricci scalar for the BSSN constrains in the same way as for the time evolution; this simplifies the generated code significantly. Don't define the (unused) upwinding differencing operators in the ADM constraints Split the BSSN constraint calculations differently into two calculations to reduce code size Add comment about constraint projection Re-generate code Update auto-generated code as generated by the current version of Kranc Update auto-generated code as generated by the current version of Kranc par: Adapt qc0-mclachlan.par to ET 05 2011, add wave extraction using ET Correct group names of imported McLachlan grid functions Regenerate code Rename tensor indices Use new Jacobian mechanism for all thorns Regenerate code Correct scheduling order Regenerate code Correct schedule, in particular for checkpoint/recovery Correct Makefile Use new fdOrder mechanism in ADM thorn Regenerate code for ADM thorn Remove (outdated) explict multi-patch version of WaveToy Initialise boundaries of A and B^i together with boundaries of Gamma^i Regenerate code Implement harmonic shift condition Regenerate code Add test case for harmonic shift Correct description of LapseAdvectionCoeff Don't replace Pi by its numeric equivalent -- Kranc does this already Remove unused function definition SetGroupTimelevels Correct definition of dot[A] advection terms Regenerate code Apply advection terms to A or B^i only if they are evolved Regenerate code Regenerate code after changes to Kranc's vectorisation Correct etaExpr to avoid (inconsequential) division by zero Regenerate code Regenerate code with current version of Kranc Some superficial changes: Handle CCZ4 in Makefile better Do not generate *_O* and _MP thorns Replace pi by Pi Regenerate code, removing the *_O* and *_MP thorns Handle Theta in helper thorn Regenerate code Update WaveToy example, generate OpenCL code as well Add thorn ML_WaveToy_CL Regenerate thorn ML_WaveToy and ML_WaveToy_CL Correct some errors in CCZ4, in particular the GammaShift parameter Regenerate code Introduce cctk_ash, retire cctk_lssh Regenerate code after Kranc update Remove TriggerGroups declaration Improve layout Add test cases for MoL/Carpet with adaptive time stepping ML_WaveToy_Test: Increase tolerance for test cases Update test results for gaussian-RK45-adaptive Regenerate code WaveToy: Don't declare Jacobian explicitly any more WaveToy: Rename local variable "k" to "kvec" WaveToyFO: Don't handle Jacobian explicitly Regenerate code after Kranc update Regenerate code after Kranc change Update OpenCL code after Kranc change
Frank Loeffler (3): Merge branches 'master', 'master', 'master', 'master' and 'master' of git://carpetcode.dyndns.org/McLachlan Merge branches 'master', 'master', 'master', 'master', 'master', 'master' and 'master' of git://carpetcode.dyndns.org/McLachlan Schedule ML_BSSN_convertToADMBaseGroupWrapper also at CCTK_POST_RECOVER_VARIABLES
Frank Löffler (1): remove outdated thornlists (closes #1156)
Ian Hinder (72): Add test cases for BSSN - 3D diagonal gauge wave ML_BSSN_Test: Remove testsuite output for constraints ML_BSSN_Test: Update testsuites to use 4th order differencing in Exact Regenerate code Remove testsuites Add new tests for ML_BSSN, ML_BSSN_O8 and ML_BSSN_MP_O8 McLachlan_BSSN.m: Apply dissipation consistently Regenerate code Regenerate testsuite McLachlan_BSSN.m: Split dissipation into a separate calculation Regenerate code McLachlan_BSSN.m: Split advection terms into a separate calculation Regenerate code Add test for individual RHS variables computed on initial data Regenerate code after Kranc ToReal change Regenerate code after OutputForm change in Kranc Regenerate code after tensor index fix to Kranc Regenerate test suite Regenerate code McLachlan_BSSN.m: Change from old TensorTools to new Kranc Jacobians for MP Regenerate code McLachlan_BSSN.m: Add InitGamma calculation Regenerate code Add poison to all tests Revert recent commits Revert "Rename tensor indices" Undo "Revert recent commits" Add missing poison in test Don't terminate on NaNChecker in tests Initialise dtlapse and dtshift in tests Set RHS variables to zero in ANALYSIS before computing them Regenerate code Regenerate testsuites with new initial gauge Rename tensor indices Regenerate code Regenerate ML_ADMQuantities and ML_WaveToyFO McLachlan_ADM.m: Add explicit "*" to quiet warning Improve finite differencing in tests Regenerate test results McLachlan_BSSN.m: Make finite difference order a run-time parameter Regenerate code Remove debugging code Use kranc script instead of calling Mathematica directly Regenerate code McLachlan_ADMConstraints.m: Fix index errors Regenerate ML_ADMConstraints McLachlan_ADM.m: Fix index errors in constraint computation Regenerate ML_ADM Fix inconsistent index positions Regenerate code (CCTK_LOOP) Makefile: Only copy back CCZ4 if it was generated ML_BSSN_Helper: Correct background value of phi in NewRad boundary condition Regenerate ML_BSSN_Helper and ML_CCZ4_Helper Add test case for NewRad boundary condition for ML_BSSN Regenerate code Use CCTK_REAL_VEC instead of ptrdiff_t for dirN variables Set apply_dissipation = "always" in test cases Regenerate code (Kranc commit d75251ed3fdfced898bbd1c9c8e940c15896640f) Regenerate code (Kranc commit b2315b27195e1b18786b80c6b9a22c3502948d77) Regenerate code (Kranc commit 17deb8b4ff45a8a3cd1645cb626f6d33a5e17beb) Regenerate code (Kranc commit b0019939fb9ef98a05431967a4cc2139f0f2a1e4) Revert "Remove tests for _MP and _O8 thorns which no longer exist." Adapt O8 and MP test parameter files to use ML_BSSN thorn and update ML_BSSN_MP_O8_bh.par: Add poison of 0 ML_BSSN_MP_O8.par: Change n_angular to number of cells from number of points test.ccl: Increase absolute tolerance of ML_BSSN_MP_O8_bh to 1e-10 McLachlan_BSSN.m: Add comment to explain confusing situation with advection term for \tilde \Gamma^i in \partial_t B^i equation Regenerate code (Kranc commit 519f2cd182c4828e74429e883139fbb8b1a11d3c) SetGroupTags.c: Ignore error if Prolongation table entry is already present Regenerate ML_BSSN and ML_CCZ4 (Kranc commit 519f2cd182c4828e74429e883139fbb8b1a11d3c) Regenerate code (Kranc commit 519f2cd182c4828e74429e883139fbb8b1a11d3c) Add ML_BSSN_sgw3d_harmonic_phi test
Peter Diener (55): Some local fixes to test Merge branch 'HEAD' of ssh://relativity.phys.lsu.edu/home/perturbed/gitroot/McLachlan Constraints Merge branch 'master' of ssh://relativity.phys.lsu.edu/home/perturbed/gitroot/McLachlan Fix RHS for phi Fix trK RHS. Fix trK RHS. Merge branch 'HEAD' of ssh://relativity.phys.lsu.edu/home/perturbed/gitroot/McLachlan Fix RHS for gamma Optimization. Replace Xt with Xtn when it appears without derivatives. Dummy commit. Merge branch 'HEAD' of ssh://relativity.phys.lsu.edu/home/perturbed/gitroot/McLachlan Beginning of matter support. Merge branch 'HEAD' of ssh://relativity.phys.lsu.edu/home/perturbed/gitroot/McLachlan Added matter terms. Fix scheduling of matter routines. Fix to Makefile. Merge branch 'HEAD' of ssh://relativity.phys.lsu.edu/home/perturbed/gitroot/McLachlan Gauges Merge branch 'HEAD' of ssh://relativity.phys.lsu.edu/home/perturbed/gitroot/McLachlan Fix convertToADMBaseCalcBSSN Schedule ML_BSSN_convertToADMBase everywhere. Head on parameter file Don't enforce lapse>0. Merge commit 'origin/master' No real change Merge commit 'origin/master' Merge commit 'origin/master' dummy merge Merge commit 'origin/master' 6th order version, upwinding version and upwinding, W-method version. Merge commit 'origin/master' Merge commit 'origin/master' Merge after running script Merge commit 'origin/master' Introduce shift advection upwinding. Update auto generated code. Undo changes to the makefile. Update auto generated files. Merge branch 'HEAD' of ssh://relativity.phys.lsu.edu/home/perturbed/gitroot/McLachlan Add support for the W-method. Add the missing helper prototype files. Merge branch 'HEAD' of carpetgit@carpetcode.dyndns.org:McLachlan.git Merge with Erik's latest changes. Clean up. Make BetaDriver spatially varying. Make sure BetaDriver routines are scheduled appropriately. Merge branch 'HEAD' of carpetgit@carpetcode.dyndns.org:McLachlan.git Fix and clarify parameter conversions. Move RHS evaluations around to optimize for 32Kb instruction cache machines. Reintroduce an optimization. Change the dafault value of apply_dissipation to "never". Fix the sign for 2nd and 6th order dissipation operators. Regenerate code.
-----------------------------------------------------------------------
hooks/post-receive -- McLachlan