v0.6.0-rc.1

This pre-release will be used for running the Geant4 GPU benchmark problems.

What's Changed

• Store edge bounding boxes on BIH and refactor for upcoming intersection capabilities by @elliottbiondo in #1478
• Improve logger when using either CUDA or MPI by @sethrj in #1482
• Check VecGeom BVH device pointers at setup and run time by @sethrj in #1481
• Implement optical Rayleigh model by @hhollenb in #1469
• Fix unnecessary linking when using VecGeom+CUDA by @pcanal in #1487
• Fix dBVH memcpy size by @esseivaju in #1490
• Fix/standardize multi-level.gdml file and improve debug logging by @sethrj in #1493
• Use canonical loop form with OpenMP by @amandalund in #1492
• Allow core and optical loops to have different numbers of track slots by @amandalund in #1491
• Add muon decay interactor by @stognini in #1456
• Fix linking Athena with VecGeom+CUDA by @pcanal in #1489
• Reconstruct "touchable" using navigation hierarchy by @sethrj in #1495
• Work around RDC-related link errors when +cuda+vecgeom~shared by @esseivaju in #1498
• Move alongstep directory to top level by @sethrj in #1499
• Fix redundant move warning by @esseivaju in #1500
• Fix documentation by @sethrj in #1502
• Default SD integration to "true" by @sethrj in #1501
• Add debugging information for app tests by @sethrj in #1504
• Refactor muon bremsstrahlung angular distribution by @amandalund in #1509
• Add FourVector helper functions by @sethrj in #1510
• Define hypotenuse function using fma by @sethrj in #1513
• Add a CI job for c++20 by @esseivaju in #1511
• Fix celer-sim MC truth ROOT output file generation during tests by @stognini in #1515
• Split CeleritasUtils cmake and use VecGeom RDC if available by @sethrj in #1514
• Hotfix: Disable ROCm due to now-unsupported CI OS image by @sethrj in #1521
• Update post-PR workflow to actions/download-artifact@v4 by @esseivaju in #1523
• Add a helper class for inverse transform sampling by @amandalund in #1525
• Require clang-tidy in CI by @esseivaju in #1524
• Stop CMake from cleaning root test data by @hhollenb in #1516
• Trim export data for unit tests by @sethrj in #1527
• Add CITATION file with CHEP 2024 article by @sethrj in #1529
• Add option to use spline-interpolated energy loss by @lebuller in #1496
• Add an Optical WLS by @whokion in #1507
• Rename Cerenkov to Cherenkov by @amandalund in #1533
• Add muon pair production by @amandalund in #1518
• Update Readme and Contributing doc by @pcanal in #1532
• Tidy headers by @esseivaju in #1534
• Fix hepmc3/macro initialization in celer-g4 by @sethrj in #1535
• Clean up optical mock test data by @hhollenb in #1519
• Run clang-tidy on modified files only by @esseivaju in #1531
• Store a single process-integrated energy loss and range table per particle by @amandalund in #1536
• Add piecewise integrator and CDF utilities by @sethrj in #1537
• Fix clang-tidy weekly cron by @esseivaju in #1541
• Extend BIH capabilities for intersection operations by @elliottbiondo in #1479
• Remove curand performance test by @sethrj in #1543
• Add along-step and tracking cut optical executors by @sethrj in #1540
• Add missing shell setup in README by @pcanal in #1545
• Enable optical stepping loop by @sethrj in #1546
• Simplify UnitProto after building a CsgUnit by @esseivaju in #1415
• Implement device bitset by @esseivaju in #1547
• Build VecGeom in weekly clang-tidy cron by @esseivaju in #1551
• Document and slightly refactor Bethe-Heitler interactor by @sethrj in #1553
• Define precision-agnostic Constant class by @sethrj in #1549
• Adapt bitset word size by @esseivaju in #1554
• Fix accel examples and related CI issues by @sethrj in #1557
• Update copyrights by @sethrj in #1560
• Prepare to update copyrights by @sethrj in #1559
• Update docker and Spack CI by @sethrj in #1555
• Enable more dependencies in clang-tidy checks by @esseivaju in #1561
• Reduce celer-sim memory usage and improve its CELER_LOG transport output by @stognini in #1550
• Support sense caching for future lazy evaluation by @esseivaju in #1539
• Support building infix logic representation by @esseivaju in #1530
• Update CMake policy by @sethrj in #1563
• Add 'LogicalTrue' functor by @sethrj in #1564
• Fix Doxygen member group formatting by @amandalund in #1567
• Fix unnecessary use of inline constexpr by @sethrj in #1568
• Replace "volid" with "vol_id" throughout ORANGE by @elliottbiondo in #1486
• Compose caching functionality of LazySenseCalculator by @esseivaju in #1569
• Update expected safeties in Vecgeom tests and CMake version range by @mrguilima in #1552
• Add support for Geant4 navigator through offload interface by @sethrj in #1558
• Hotfix: fetch tags manually to fix CI by @sethrj in #1580
• Fix spurious vecgeom warning by @sethrj in #1584
• Import and use muon+hadron EM physics parameters by @amandalund in #1581
• Update Perlmutter env by @esseivaju in #1587
• Add ARM runner to CI by @sethrj in #1585
• Require less ancient CUDA architecture and OpenMP implementation by @sethrj in #1578
• Fix weird edge cases when trying to find_package(Celeritas) by @sethrj in #1586
• Propagate code citations into user manual by @sethrj in #1575
• Extend Coulomb scattering to support additional particle types by @amandalund in #1574
• Use lazy sense calculator by @esseivaju in #1576
• Release v0.5.1 by @sethrj in #1577
• Improve and fix documentation by @sethrj in #1588
• Improve diagnostic output when offloading by @sethrj in #1589
• Add optical model importer and refactor imported optical materials by @hhollenb in #1520
• Add example offloading template by @stognini in #1573
• Output diagnostic information about core sizes by @sethrj in #1591
• Change PrimaryGeneratorAction to be a wrapper by @sethrj in #1593
• Simplify and add to system config diagnostics by @sethrj in #1590
• Add global Celeritas input definition by @sethrj in #1562
• Set step length and other data in reconstructed tracks by @sethrj in #1595
• Refactor primary generator using new inp by @sethrj in #1583
• Add hit manager output diagnostic by @sethrj in #1596
• Update core advisors by @sethrj in #1598
• Always select at-rest process for stopped tracks by @amandalund in #1592
• Add postfix to infix conversion utility by @esseivaju in #1582
• Write out optical size diagnostic by @sethrj in #1597
• Move physics lists out of details and rename tracking offload by @sethrj in #1603
• Use BIH to accelerate background_intersect by @elliottbiondo in #1599
• Update perfetto to v49.0 by @esseivaju in #1608
• Use external G4VG by @sethrj in #1605
• Add TrackingManagerConstructor and SharedParams::GetMode by @sethrj in #1606
• Move PolyEvaluator to corecel/math by @amandalund in #1610
• Add JLSE CMake preset and a few small fixes by @amandalund in #1611
• Add ultra-simple tracking manager integration by @sethrj in #1609
• Fix multiple release() calls to ensure nullptr is not used in process construction by @amandalund in #1612
• Rename fast simulation offload by @sethrj in #1613
• Implement cubic spline interpolation by @amandalund in #1607
• Rename Generic grid to Nonuniform by @amandalund in #1616
• Add a calculator to integrate range from energy loss by @amandalund in #1544
• Add G4Ellipsoid without bottom/top cuts by @elliottbiondo in #1617
• Fix CUDA+VecGeom linking with Ubuntu by @pcanal in #1618
• Add comment directing to CMake doc of feature being used by @pcanal in #1619
• Rename EXTERNAL CMake variable by @sethrj in #1621
• Add safety visualization tool by @sethrj in #1614
• Remove pointers from Geant4 labels by @sethrj in #1620
• Support different grid spacings for cross section data by @amandalund in #1622
• Add high-level user/fast sim integration helpers by @sethrj in #1615
• Support spline interpolation on a nonuniform grid by @amandalund in #1625
• Fix crash in multithreaded celer-g4 by @amandalund in #1627
• Prune primaries that start outside the world volume by @sethrj in #1624
• Fix loading of GDML files with reflection by @sethrj in #1626
• Add generic geometry test interface to reduce duplication by @sethrj in #1623
• Bump VecGeom minimum version to 1.2.10 by @pcanal in #1628
• Fix navigation and mapping of reflecting volumes by @sethrj in #1629
• Construct CoreParams from new problem input in celer-sim by @sethrj in #1601
• Flush perfetto TrackEvents by @esseivaju in #1630
• Move Geant4 SD integration into celeritas/ext by @sethrj in #1631
• Add perfetto track counters to accel LocalTransporter by @esseivaju in #1634
• Fix reading and writing HepMC3 events by @amandalund in #1635
• Improve testing and documentation for VecGeom by @sethrj in #1639
• Fix rendering of README by @dalg24 in #1640
• Add support for Z cuts when converting G4Solids by @elliottbiondo in #1638
• Update externals and enforce G4VG version by @sethrj in #1641
• Add ORCID to citation by @esseivaju in #1645
• Increase ccache size for asanlite job by @sethrj in #1646
• Update cmake-git-version by @sethrj in #1643
• Fix getenv_flag by @amandalund in #1647
• Refactor log handlers for better reuse by @sethrj in #1636
• Support sensitive detectors in replica/parameterized volumes by @sethrj in #1649
• Hotfix: remove Ubuntu 20 from CI configuration by @sethrj in #1651
• Add an option to enable status checker through celer-sim by @sethrj in #1650
• Improve geometry testing by @sethrj in #1652
• Add support for elliptical cylinders and elliptical cones in orange and g4org by @elliottbiondo in #1642
• Hotfix: prevent vecgeom 2.0 in regular builds by @sethrj in #1655
• Support post-step touchable location and step status by @sethrj in #1653
• Support surface-based VecGeom 2.x navigator by @mrguilima in #1422
• Approximate G4Torus using inner/outer bounding cylinders by @elliottbiondo in #1654
• Rename CoreTrackView methods by @amandalund in #1658
• Refactor geometry tests into common file by @sethrj in #1656
• Use extern static data for versions, configuration by @sethrj in #1657
• Support volume-specific uniform field by @amandalund in #1659
• Fix link errors with CUDA RDC and --as-needed flag by @pcanal in #1661
• Use CLI11 for front end command line parsing by @sethrj in #1660
• Move offload template into a celeritas::example namespace by @sethrj in #1663
• Implement discrete optical physics by @hhollenb in #1604
• Update frontier toolchain for Feb 2025 by @sethrj in #1668
• Remove 'default_stream' option by @sethrj in #1667
• Fix downstream RDC linking and add versioning by @pcanal in #1665
• Extend CudaRdc compatibility by @pcanal in #1671
• Default downstream CUDA/HIP architectures by @sethrj in #1669
• Hide Thrust from .cc code by @sethrj in #1670
• Add an example user action for gathering statistics on GPU by @sethrj in #1664
• Initialize MPI in high-level integration classes by @sethrj in #1672
• Add quantity support to FindInterp and template Turn by @sethrj in #1675
• By default load CudaRdcUtils and print config message by @sethrj in #1676
• Support VecGeom 2.0 by @sethrj in #1674
• Fix bremsstrahlung process construction when using only a single model by @amandalund in #1677
• Add 3D Cylindrical field map by @esseivaju in #1662
• Lazily fetch volumes for UniformAlongStepFactory by @rahmans1 in #1666
• Remove G4EmExtraPhysics from FtfpBertPhysicsList by @whokion in #1644
• Define dependency helper targets and reorganize CMakeLists by @sethrj in #1673
• Update CudaRdcUtils to v3 by @pcanal in #1679
• Add Celeritas cmake command wrappers for downstream code by @sethrj in #1678
• Fix static builds and CI reporting by @sethrj in #1680

New Contributors

• @rahmans1 made their first contribution in #1666

Full Changelog: v0.5.1...v0.6.0-rc.1