Changelog

The format is based on Keep a Changelog and this project (mostly) adheres to Calendar Versioning.

All major changes to the Athena++ private repository between each version/tag are summarized in this CHANGLEOG.md document. Each version has an Issues and Pull Requests section, whose subsections are automatically populated from the issue/PR labels. The list entries contain links to the private repository issue tracker #N id.

Additionally, the changes are manually summarized using the following categories:

Added: for brand new features or extended capabilities
Fixed/Changed: for bug fixes or modified behavior
Removed: for removed functionality

The automatically-generated content should be used for reference when writing these sections. At this time, both the private and public GitHub Release Notes are started by copy/pasting from these sections.

Unreleased

v24.0 (2024-04-01)

Implemented enhancements:

Fixes athena_read's unnecessarily large memory allocation #587 (dgagnier)
Divergence-preserving prolongation of coarse/fine shared face-centered fields for 3D spherical polar and cylindrical grids #584 (mariareneemf)
Divergence-preserving prolongation of coarse/fine shared face-centered fields for 3D spherical polar grids #582 (mariareneemf)
Remove unused DEBUG macro #531
Make DEBUG macro a binary 1/0 instead of string #533 (felker)
Update cpplint.py #524 (felker)
Add Jenkins script on stellar #517 (changgoo)
Chemistry update units #489 (changgoo)
Updated Python scripts #471 (c-white)
Make turbulent velocity field resolution independent #461 (sanghyukmoon)
Added error message for rotating frame in 1d spherical polar #424 (msbc)
Clarify units in EOS input file #423 (msbc)
Add hydro and MHD regression tests for quirk.cpp #386 (felker)
Prepare to open-source the development repository #361 (felker)
Fix the switch to default RKL2 STS #350 (pdmullen)
Minor update of shearing box #348 (tomo-ono)

Fixed bugs:

Timestep is incorrect for some GR metrics #545
Is this pdata->pnext supposed to be pnew->pnext here? #528
AMR in GR Does Not Work (Undefined Behavior) #475
Small bug in specifying uov output #436
MPI disk simulations hang on Berkeley's Savio cluster #430
Wrong density used for relativistic passive scalar prolongation #419
h5py 3.x releases causes issues with plot_slice.py #393
Numerical artifacts in the gr_torus problem #381
Shear flag is set to false for all corner neighbors #377
HLLC can get NaNs #352
4th order reconstruction and outflow BCs bug #298
Add more runtime NGHOST checks (especially for SMR/AMR) #163
GR-AMR incompatibility #64
Remove division-by-zero bug for viscosity + 3d polar boundary condition #581 (apbailey)
Replace more nonconstant references with const #577 (felker)
Fix x1v for Curvilinear+PPM #566 (forrestglines)
Remove all possibly extraneous DeleteAthenaArray calls #560 (felker)
Fixed Typo in Error Messages #557 (Rgokemeijer)
Miscellaneous small fixes #549 (c-white)
Add trap for MHD with xorder=4,4c selections #535 (felker)
Suppress 2x ICX default warnings #532 (felker)
Stop accessing a deleted pointer #529 (felker)
Fix a potential bug in the characteristic projection #497 (bcaddy)
Fix an indexing bug in x3v in disk.cpp #458 (tomo-ono)
Change SMALL_NUMBER to 10^-4 in the HLLD-series Riemann solvers #457 (tomidakn)
Fix hang when using shearing box + STS #456 (pdmullen)
Apply output time cadence changes even if no input file is provided during restarts #454 (felker)
Fixing the restart error reported in #449 #453 (tomidakn)
Fixed bug in fixup routine for SRMHD #452 (c-white)
Fixed passive scalar + AMR bug #446 (c-white)
Initialize pointer fv_new_ to nullptr #445 (kahoooo)
Fix bug specifying uov output #440 (kahoooo)
Fix dependency of passive scalar boundary communication #434 (tomo-ono)
Fixed typo in gr_torus problem generator #431 (c-white)
Tiny fix for reading Schwarzschild .athdf files #425 (c-white)
Corrected density used in relativistic passive scalar prolongation #420 (c-white)
Extend the x2 loop limit by 1 for x3 outflow outer bc #404 (felker)
Update plotting slices and add polar projection view #394 (kurtsansom)
Fixed typo in GR Bondi problem generator #385 (c-white)
Fix of bug in the strat pgen for a restart run #384 (tomo-ono)
Fix typo in setting of shearing boundary flag #379 (tomo-ono)
Fix shearing boundary flag for corner neighbors #378 (tomo-ono)
Fix a bug in OA hydro source term #368 (tomo-ono)
Correct calculations of the energy viscous flux #363 (tomo-ono)
Fix a bug in history output again #360 (kahoooo)
Fix a bug that takes up unnecessary space in restart file #358 (kahoooo)
Fix a bug in history output #357 (kahoooo)
Make variables used in simd loops private #355 (pgrete)
Don't load conserved variables into additional buffers from .rst #353 (felker)
Fix 4th-order hydro with non-periodic BCs #347 (pdmullen)
Reduce overly greedy NGHOST requirement for PPM+MHD #346 (felker)
Bug fix of OA for MHD #345 (tomo-ono)

Closed issues:

athena_read allocates unnecessarily large amount of memory #586
Array indexing errors in llf_mhd_rel_no_transform.cpp #568
Access ParameterInput instance from user boundary conditions #563
Remove extraneous DeleteAthenaArray() calls #559
Inconsistency in documentation of user-defined source terms? #556
Resurrect Princeton Jenkins server #515
athena_read.py produces corrugated spherical-polar x2v coordinates due to float precision at high resolution #498
FileNotFoundError for regression tests when "python" is not in PATH but "python3" is #479
Command-line output#/dt and output#/next_time are handled incorrectly #451
Method for gravity source terms should be documented #450
Restart with too few cores produces confusing error #449
pgen 'disk' behaving strangely on KNL with --eos adiabatic #435
Unable to restart a simulation #429
HDF5 writer: Segfault in memory access #416
Incorrect running index for face-centered variable #403
Write generated configuration to file #396
Possiblly wrong code in EquationOfState::ApplyPrimitiveFloors #395
linwave: initial conditions do not scale with resolution #392
linwave1d with iso_sound_speed != 1 #391
Passive scalars behaving incorrectly upon refinement/de-refinement #390
Output synchronization: Unable to generate same number of output files for two simulations #387
Viscous diffusion in rotating frame #373
Unwanted spiral density waves in parallel sims #370
Potentially Ill-Defined Ambipolar Diffusion Constant Diffusivity #369
Unsafe SIMD loop(s?) in WeightedAve #349
Is the restriction of NGHOST for MHD with the 3rd order reconstruction scheme necessary? #339
Output next time is ignored on startup #310
Passive scalar issue with AMR #365

Merged pull requests:

Remove Travis CI YAML and badge from README #592 (felker)
Add better exclusion for MHD+shearing box+SMR with refinement in x3 dir at shearing boundary #591 (felker)
Fully-implicit cosmic-ray diffusion module using Multigrid #550 (tomidakn)
Replace std::fabs with std::abs #523 (felker)
Trigger PR test on new Jenkins server and fix all violations #518 (changgoo)
Add chemistry module built on passive scalars module #496 (munan)
Add non-relativistic radiation transport solvers (explicit and implicit) and cosmic ray transport #492 (felker)
Chemistry update #491 (doraemonho)
Multigrid Gravity Solver Release Candidate #484 (tomidakn)
Do not hardcode "python" executable name #480 (felker)
Rename source term functions #437 (tomo-ono)
Added "dcycle" option to output data at constant cycle intervals #417 (tomidakn)
Support cherrypicking scalars #414 (kahoooo)
Allow cherrypicking variables to be output #413 (kahoooo)
Low-dissipation HLLC/HLLD solvers (Minoshima et al. 2021) #380 (tomidakn)

v21.0 (2021-01-06)

Removed

Multigrid solver and related regression tests (removed directly from v21.0-dev)

v21.0-dev (2021-01-06)

Implemented enhancements:

Need more flexible task control in STS #270
Enable AMR/SMR and source terms with -sts #250
Robustness improvements to relativity #316 (c-white)
Change primitive velocities in SR #303 (c-white)
Update passive scalars to have user source terms and work with relativity #297 (c-white)

Fixed bugs:

Spurious waves emerging from corners using the intel 19.1 compiler and MPI? #340
EOS Riemann regression test fails #335
Unused and uninitialized variables with orbital advection #332
regression tests for diffusion do not work & failed #328
Reconstruction of passive scalars uses wrong array size #306
malloc error with gcc compiler, spherical polar coordinate and xorder=3 #304
RK3 integrator breaks conservation #300
Runtime error "pure virtual method called" with intel compiler #289
Test GR/SR compatibility with passive scalars #281
New Limitation of MeshBlock Size #343 (tomo-ono)
Move Poisson solve to after ptlist->DoTaskListOneStage() #309 (pdmullen)
Add support for NSCALARS > NWAVE #307 (msbc)
Fix logic in bvars_sts designation #305 (pdmullen)

Closed issues:

bugs in using std::min or std::max for single precision #326
Mismatch of the function argument in WeightedAve for FaceField #318

Merged pull requests:

Implemented the source term formula for gravitational acceleration #344 (tomidakn)
Enable STS in Shearing Box with Orbital Advection #342 (tomo-ono)
Logical location (long)is used to set array index (int) -- Jenkins passed! #341 (changgoo)
Fix lack of deallocating dynamic memory #338 (tomo-ono)
fix dependency in tasklist #337 (tomo-ono)
Increased thresholds for eos_riemann test #336 (msbc)
fix EMF flux correction in shearing box #334 (tomo-ono)
fix unused and uninitialized variables #333 (tomo-ono)
Fix conditionals in TimeIntegratorTaskList fns for compatibility with STS #331 (pdmullen)
Docstring Style Correction #330 (changgoo)
Correct the input block name in the turbulence regression test and make driving isotropic regardless of the box shape #329 (changgoo)
put cast in the min or max function #327 (tomo-ono)
Implement a New Discretization of the Gravitational Stress Tensor #325 (pdmullen)
Change in turbulence input style #324 (changgoo)
Fix Bug with Mesh Structure Output #322 (tomo-ono)
Orbital Advection & New Shearing Box #321 (tomo-ono)
Fix typo: bounday_flag.cpp to boundary_flag.cpp #320 (changgoo)
Fix mismatch between WeightedAve prototype and definition #319 (pdmullen)
Improve performance of turbulence driver #317 (changgoo)
Update to bug fix on fft pgen #315 (changgoo)
Fix typo in HLLC and bug in Hydrogen EOS #314 (msbc)
Bug fixes for FFT; New regression test #313 (changgoo)
Modifying the KH problem and a input file for MHD KH #312 (tomo-ono)

v19.0 (2019-08-06)

Removed

Multigrid solver and related regression tests (removed directly from v19.0-dev)

v19.0-dev (2019-08-06)

Switched to using CalVer (Calendar Versioning) scheme from SemVer (Semantic Versioning). v1.1.1-dev was the previous tag and version.

Implemented enhancements:

Extend user-defined history outputs beyond parallel summed quantities #237
Add test coverage for Roe Riemann solvers #231
Add "logging" module to all regression tests to replace calls to print() #203
Warn users about proper polar vs. polar_wedge boundary condition flag #196
Loading SMR data at native resolution #173
Create cleaner failure mode for incorrect input parameter file text formatting #168
Extend vis/python/uniform.py to 2D HDF5 files #164
(Non-Zero) Mean Density Must Be Set for FFT Self-Gravity with Periodic BC's #155
include input parameters in hdf5 output? #143
Extend TaskList to support more than 64 task IDs #286 (tomidakn)
Add timing to regression tests #284 (msbc)
Throw error when reading HDF5 files with ghost zones if their number hasn't been specified. #280 (msbc)
Add user ability to specify names of libraries when linking #277 (c-white)
Remove MeshBlock dependency of FFTDriver #276 (changgoo)
Move MGGravity from GravityDriver to MeshBlock #275 (tomidakn)
Add input option to output more timestep diagnostics #273 (felker)
Refactor MeshBlockTree to save memory footprint #262 (tomidakn)
Replace AthenaArray<T>::InitWithShallowCopy with C++ references #261 (felker)
Rewrite shearing box capabilities #260 (felker)
Remove do-nothing DeleteAthenaArray calls #258 (felker)
Refactor bvals/ and Mesh::AdaptiveMeshRefinement() #257 (felker)
Replace AthenaFFTComplex with std::complex<Real> #256 (felker)
Deduplicate 4x STS regression tests #252 (pdmullen)
Adds logging module to all regression tests to replace calls to print() #251 (msbc)
Global initialization of turbulence realization #241 (changgoo)
Enable FFT for an entirely refined mesh #240 (changgoo)
Optimize performance of SR/GR calculations #235 (beiwang2003)
Improve all enumerated types #234 (felker)
Add test coverage for MHD and Hydro Roe Riemann solvers #233 (msbc)
Replace holdovers from C with C++11 alternatives #219 (felker)
Update ShowConfig for general EOS #218 (msbc)
Extend line plotting to .athdf files #211 (c-white)
Make C++ exception handling optional; standardize more C++11 headers and namespace usage #190 (felker)
Extend Kelvin-Helmholtz problem generator #187 (felker)
Use C++ compiler front ends for main/documented "configure.py --cxx" options but support C front ends; add built-in support for Apple+LLVM with OpenMP workaround #186 (felker)

Fixed bugs:

athena_read.py silently fails when ghost zones are saved to file #279
MPI + continuous turbulence driving with OU process is broken #267
Invalid read in AMR #229
NaN'ing outputs may not be caught by some regression tests #223
Bug with permuting fft directions in athena_fft.cpp #217
1D AMR not refining third component of magnetic field correctly #212
MPI/AMR segfaulting #199
RECV_* boundary buffer receive tasks are missing dependencies on matching INT_* integration tasks #198
Possible incorrect memory access in spherical_polar coordinates: private & public version #192
athena_read.athdf() handles ghost zones incorrectly #184
Possible race condition in OpenMP #183
SMR causes issues with cell sizes difference warning in spherical coordinates with user mesh generator #180
uniform.py now writes all 0s when overwriting a file #179
OpenMP is broken; add regression tests to prevent this in the future #175
Some regression tests never fail #174
Python reader fails for slices taken across multiple blocks #159
Python HDF5 reader no longer returns refinement levels #156
Is ApplyPrimitiveFloors() being called too often? #153
Fix bug when saving zero scalars to restart file #285 (msbc)
Fix bug when saving multiple scalars to restart file #283 (msbc)
Fix minor issues with Python HDF5 reader #278 (c-white)
Fix curvilinear reconstruction, add regression test scripts (passive scalars, MPI shearing box, etc.), and other QoL improvements #274 (felker)
Revert GR optimizations #272 (c-white)
Fix a NaN bug (#267) caused by the incorrect out-of-place FFT #268 (changgoo)
Fix bug in spherical plotting script #265 (c-white)
Fix shearing box + MPI #264 (felker)
Fix FFT permutation bug #238 (changgoo)
Improve NaN handling in Python scripts #224 (felker)
Fix fail criterion for diffusion/*_diffusion* tests #222 (pdmullen)
Fix typo in 1D prolongation of magnetic field #216 (c-white)
Fix general EOS bugs and add eos_test pgen #214 (msbc)
1D MHD AMR fix #213 (c-white)
Correct energy bug in PrimitiveToConserved in general_hydro.cpp #208 (msbc)
Apply minor fixes to Python scripts #201 (c-white)
Fix memory access issue with spherical diffusion #195 (c-white)
Improve post-processing and plotting scripts #185 (c-white)
Fix minor OpenMP error and duplicate MPI regression test for OpenMP, hybrid MPI+OpenMP #178 (felker)
Fix and extend Python scripts for analysis of HDF5 files #167 (c-white)
Fix Python HDF5 reader's ability to return refinement levels #158 (msbc)
Eliminate redundant calls to ApplyPrimitiveFloors() #154 (felker)

Closed issues:

Add optional time/sts_nstage_out input parameter #249
Deduplicate 4x STS regression test scripts in diffusion/ #242
"register" keyword used in src/fft/plimpton/ files is deprecated and was removed in C++17 #230
Make athinput a file extension as opposed to a file prefix. #226
1D MHD SMR fields wrong at refinement boundaries with more than 2 ghost zones #209
Add interface for operator splitting and super-time-stepping #172

Merged pull requests:

Add user-defined passive scalar floor option in input file #291 (munan)
Add MHD capabilities to the general EOS module #282 (msbc)
Add passive scalars #263 (felker)
Extend automatic and manual load balancing beyond AMR #259 (tomidakn)
Fix edge-case bug in typecast function. #255 (msbc)
Implement OU Process for driven turbulence and Helmholtz decomposition #244 (changgoo)
Remove "register" keyword from 2x .cpp files in plimpton/ #232 (felker)
Fix general EOS hydro function names and prepare for general EOS MHD extension #228 (msbc)
Add compatibility checks for polar, polar_wedge boundary conditions to class constructor #200 (felker)
Generalize the hydrodynamics equation of state capabilities for non-ideal gases #197 (msbc)
Revert to using 1D pencil arrays for performance #193 (tomidakn)
Update input params for diffusion regression problems #188 (pdmullen)
Integrate code coverage analysis into regression test framework and continuous integration #181 (felker)
Implement RKL1 Super-Time-Stepping #176 (pdmullen)
Add fourth-order hydrodynamics solver #157 (felker)
Add new script for plotting 2D slices #151 (c-white)
Add HDF5 reader and pgen for initializing from HDF5 #146 (c-white)

v1.1.1 (2018-07-31)

Removed

Multigrid solver and related regression tests (removed directly from v1.1.1-dev)

v1.1.1-dev (2018-07-31)

Added

Performance improvements for SR calculations
Pragmas and clauses from OpenMP 4.0 and 4.5 for SIMD-enabled functions inside vectorized loops (instead of depending on the compiler to inline the function calls)
LLF Riemann solver for Newtonian hydrodynamics
Rules for checking Python style of Athena++ scripts with flake8
Interactive plotting for spherical polar coordinates results
Several new input files and problem generators

Fixed/Changed

Re-enabled SIMD vectorization for Roe-type Riemann solvers and rewrote eigenmatrix calculations to improve performance
Improved readability of TimeIntegratorTaskList
Fixed spherical coordinates terms for non-ideal MHD
Fixed reflective symmetry preservation for hydrodynamic viscosity calculations
Eliminated small floating-point errors when analyzing uniform grid results using included Python HDF5 reader
Changed turbulence driving switches to avoid possible bug during initial cycle
Plugged MPI resource leaks in Multigrid

Issues and Pull Requests:

Fixed bugs:

Gravity boundary functions not set when enrolled #148
Viscosity terms break Rayleigh-Taylor symmetry #144
SIMD vectorization disabled for Roe-type Riemann solvers #126
Memory leak in jeans_3d.py test MPI run #115
athena_read.py athdf() results in small floating-point errors #111
Remove unused GravityBoundaryFunction_[] array from Mesh class #149 (felker)
Cleanup minor issues before v1.1.1 release; fix viscosity asymmetry #147 (felker)
Return exact floating-point values when reading HDF5 coordinates #145 (c-white)
Fixed spherical coordinates for non-ideal MHD #142 (tomidakn)
Added interactive spherical plotting: #139 (c-white)
Use #pragma omp declare simd for functions called in SIMD loops #138 (felker)
Plugged MPI resource leaks in Multigrid. #137 (tomidakn)
AMR Fix #133 (tomidakn)
Bug fix for initial turbulence driving #132 (changgoo)

Closed issues:

Python scripts don't comply with PEP 8 style; vis scripts need documentation #96

Merged pull requests:

Improve performance for all SR problems #140 (beiwang2003)
Redefine and initialize grav_mean_rho directly in Mesh class #136 (changgoo)
Upgrade MPICH and OpenMPI in Travis CI build environments #135 (felker)
Add Python style checker and fix existing violations #134 (felker)
Improve CHANGELOG, release process, and CI testing #131 (felker)

v1.1.0 (2018-05-23)

Removed

Multigrid solver and related regression tests (removed directly from v1.1.0-dev)

v1.1.0-dev (2018-05-23)

Added

Self-gravity
- FFT
- Multigrid
Shearing box
Viscosity, resistivity, and conduction
- Anisotropic viscosity is not implemented, but it can be added as user-defined viscosity function. Can copy function's source code from Athena 4.2.
Piecewise parabolic method (PPM) with time/xorder=3 runtime option
- Robust flooring of reconstructed states
- Curvilinear and nonuniform mesh terms
Characteristic variable reconstruction (PLM and PPM) with time/xorder=2c or 3c runtime option
Redesign of time-integrator to support high-order schemes
Turbulence driving
Double precision floating-point HDF5 output
Software development tools (mostly) exclusive to private repository:
- Continuous integration (Jenkins and Travis CI)
- C++ style checker
- Expanded regression test suite flexibility and code coverage
- Code Reviews, Issue/PR categories, probot automated closing of stale Issues/PRs, protected branch status for master
- Slack workspace, CONTRIBUTING.md guide, Issue and PR templates.

Fixed/Changed

Changed OpenMP setup to coarse threading over MeshBlock; thread-safe MPI now in use.
Performance optimizations: improved vectorization, decreased memory traffic, etc.
Fixed reflective symmetry preservation (exact to double precision for hydrodynamics)
Fixed all compiler warnings and ensured C++11 compliance

Issues and Pull Requests:

Implemented enhancements:

Pass MeshBlock to user defined diffusivity function #130 (jmshi)
Use native HDF5 byte order #109
Conflicting restart and input parameters for writing output #62
Add double precision option for HDF5 output #28
Improve vectorization for loops in PPM #121 (beiwang2003)
Improve symmetry-preservation of HLLD floating-point operations #120 (felker)
Fix restarted simulation calculation of next_time for writing outputs #116 (felker)
Support double precision floating-point HDF5 output #108 (felker)
Add templates for Issue and PR; creating draft of CONTRIBUTING.md #93 (felker)
Make athdf a class which will delay reading until data is requested #84 (msbc)
Improve TimeIntegratorTaskList performance #80 (felker)
Use AVX-512 registers for vectorized loops on Skylake; enable OpenMP 4.0 SIMD with newer GCC versions #77 (felker)
Require negative xrat for user-defined mesh generation #71 (felker)
Replace "#pragma simd" with "#pragma omp simd" for icc 18 #66 (felker)
Make athinput reader Python 3 friendly in athena_read.py #60 (msbc)
Add athinput reader to athena_read.py #59 (msbc)
Ensure Python 2 and 3 compatibility of regression test suite #48 (felker)
Improve Intel AVX2 vectorization of hydrodynamics routines #44 (felker)

Fixed bugs:

Regression tests are writing .pyc files #104
Need to address g++, clang++, icc compiler warnings #94
PPM on spherical grid causes segmentation fault #76
Summed outputs have significant errors #73
Multigrid gravity NGHOST > 2 causes artifacts #67
FFTBlock.norm_factor_ may be uninitialized #51
Regression test suite with MPI causes nonblocking IO write errors on macOS #47
UserWorkAfterLoop is called before final outputs #45
Restart regression test fails due to FFT input parameter parsing #42
Incorrect initialization of scopy_ used for AthenaArray deep copy #38
Need better default behavior when setting start_time #36
Fix SIMD vectorization and regression tests for diffusion module #127 (jmshi)
Use simple average when computing cell-centered B #122 (felker)
Fix floating-point asymmetry of AMR cell-centered prolong/restrict operators #114 (felker)
Fix asymmetry of Coordinates x1f for multiple MeshBlock grids #112 (felker)
Fix general asymmetry of x1f, dx1v Coordinates terms; ensure exact float64 symmetry of PPM stencils #98 (felker)
Implement various PPM fixes #79 (felker)
Fix to self-gravity FFT and MG Solve #72 (pdmullen)
Add missing space in hdf5_path string to fix bug #70 (msbc)
Fix self-gravity energy source term #63 (pdmullen)
Initialize FFTBlock::norm_factor_ to 1 #57 (msbc)
Improve stability of FFT gravity solver #40 (changgoo)
Reconstruction for energy shouldn't be done for the isothermal EOS. #37 (changgoo)

Closed issues:

C++ source code does not conform to Athena++ Style Guide #95
Wiki page on OpenMP parallelization is out of date #92
Require code reviews before merging pull requests #88
Create Slack workspace for Athena++ #61
Allow the regression test suite to accept platform-specific compiler and execution options #55
Regression test suite incompatible with Python 3 #46
Add optional runtime parameter for reducing/suppressing per-cycle summary info #39
Standardize sqrt() and cbrt() calls for C++11 compliance and portability #27

Merged pull requests:

Fix Bash error handling of C++ style checker #125 (felker)
Remove inverse h_coeff from Coordinates class #124 (jmshi)
Add diffusion physics capabilities #123 (jmshi)
Use default simdlen for a memory intensive loop in PLM #119 (beiwang2003)
Use loop fusion and define simdlen(SIMD_WIDTH) for PLM #118 (beiwang2003)
Add default implementations of Laplacian operators to Coordinates class #117 (felker)
Codify and fix C++11 standard compliance #113 (felker)
Improve HLLD vectorization #110 (beiwang2003)
Add Google C++ Style Guide linter and check with CI #107 (felker)
Fix main GCC compiler warnings; strictly test code changes for new warnings #105 (felker)
Use C++ style static_cast operator instead of C style type casting #101 (felker)
Add continuous integration with Travis CI and Jenkins #97 (felker)
Add Bash wrapper to join_vtk++ for many MeshBlocks, output steps #91 (felker)
Add CODEOWNERS file to new .github/ root dir #90 (felker)
Add ShearingBox component #89 (jmshi)
Improve floating-point precision consistency throughout code #82 (felker)
Add primitive variable flooring functions to the EquationOfState class #81 (felker)
Add MG and FFT convergence regression tests based on 3D linear Jeans instability #75 (alwinm)
Add cylindrical and spherical polar coordinates limiter terms to PPM #69 (felker)
Redesign the TimeIntegratorTask list to enable time integration methods with order, nsub_steps, nregisters > 2 #49 (felker)
Enable nonuniform reconstruction with Mignone PPM limiter #43 (felker)
Add ncycle_out optional input parameter #41 (felker)
Add FFT gravity solver #35 (changgoo)

v1.0.1 (2017-08-22)

Added

Scripts and documentation for releasing from private to public repository

Fixed/Changed

Upgraded left and right reconstructed primitive state AthenaArray in Hydro class to 3D (from x1-sliced 1D arrays)
Riemann solvers now directly fill electric field arrays

Issues and Pull Requests:

Implemented enhancements:

Add code structure diagrams to Wiki #34

Fixed bugs:

Alfven wave test fails due to incorrect error thresholds #29
Cannot compile properly on tiger #8
Do not access de-allocated memory in SliceOutputData #32 (dradice)

v1.0.0-beta (2016-11-28)

Added

BSD license

Fixed/Changed

Redesigned Coordinates class

Issues and Pull Requests:

Implemented enhancements:

ParameterInput should be generally accessible #20
Need OMP-ready scratch array for relativistic MHD Riemann solver #6
Small configure script enhancements #21 (jzuhone)

Fixed bugs:

Blast problem missing header needed to compile #25
MPI + HDF5 writer with ghost zone output causes segmentation fault #23
Compilation failure when using HDF5 without MPI #22
1D Newtonian MHD broken with periodic boundary #16
Problem with fields at physical boundaries in 1D #15
Line missing in src/pgen/jet.cpp #12
Compilation error -- accessing invalid member of RegionSize #11

Closed issues:

Add boundary functions based on primitive variables #17
Need flux correction for source terms #9

v0.3.0 (2015-08-31)

Added

Fixed/Changed

Issues and Pull Requests:

Fixed bugs:

Uninitialized variable root_level is used #10

v0.2.0 (2015-04-10)

Added

Fixed/Changed

Issues and Pull Requests:

Fixed bugs:

Restart output problem due to next_time=0 #7

Closed issues:

Add MeshBlock boundary condition for MHD #4

v0.1.0 (2015-02-02)

Added

Fixed/Changed

Issues and Pull Requests:

Fixed bugs:

Problem with boundary functions in 1D hydro shock tube #5

Closed issues:

Add C++ style document to wiki #1

* This Change Log was automatically generated by github_changelog_generator