lars/geant4
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Import Geant4 10.3.0.beta source tree

Browse Source
This commit is contained in:
Gabriele Cosmo
2016-06-30 14:12:05 +02:00
parent a654a7ab1f
commit 4ec577e5c4
2021 changed files with 100995 additions and 78277 deletions
Download Patch File Download Diff File Expand all files Collapse all files
ReleaseNotes/Beta4.10.3-1.txt
+770
View File
@@ -0,0 +1,770 @@
Geant4 10.3-beta-01 Release Notes
---------------------------------
30 June 2016
Migration Notes & Main New Features
-----------------------------------
o Introduced new utility class G4MTBarrier implementing a synchronization
point between threads.
o Added handling of multiple actions to be instantiated and handled by
user code. Added G4MultiRunAction, G4MultiPrimaryAction and
G4Multi{Stepping,Tracking}Action.
o Added ability to scale shapes' dimension in X, Y and Z.
Added possibility to import/export scaled shapes in GDML.
o Updated properties of particles according to PDG-2015.
o Introducing floating level base to G4Ions and related classes.
o Disabled multiple-scattering at post-step for position change,
now all happening along-step.
o All EM parameters now handled by G4EmParameters class and can be modified
via UI commands, G4EmProcessOptions class is obsolete and will be
removed soon.
o Major revision of the QGS final-state model (note: this was already
included in 10.2.beta, but did not enter in 10.2 due to the missing
re-tuning of parameters): implemented Reggeon cascading and "Fermi motion";
new algorithm for the determination of the kinematical properties of
partons; improved formation of the residual nucleus.
o New tuning of the FTF model parameters; introduced smearing of delta-isobar
mass and improved di-quark fragmentation into baryons in Lund string
fragmentation.
o Introduced drawing and filtering by touched volume in visualization.
o New data sets G4ENSDFSTATE-1.3.1, G4RadioactiveDecay-4.4.1 and G4TENDL-1.2.
----------------------------------------------------------------------------
Technical Notes
---------------
o Tested platforms:
+ Linux, gcc-4.8.4.
Tested on 64 bit architectures (Intel or AMD) with CERN CentOS
Linux 7 (CC7) (based on CentOS Linux 7).
Versions of Geant4 have also been compiled successfully on other
Linux distributions, Ubuntu, Debian, Suse or other RedHat systems.
+ MacOSX 10.11, clang-3.7 (Apple LLVM/Clang-7.3.0)
* For multi-threading on MacOSX, -clang- compiler has to be used.
+ Windows/10 with Visual C++ 14.0 (Visual Studio 2015)
* Multi-threading mode currently -not- supported on Windows.
o More verified configurations:
+ Linux, gcc-4.9.4, gcc-5.3.0, gcc-6.1.0
+ Linux, Intel-icc 15.0, 16.0.
+ MacOSX 10.10 with clang-3.6
+ MacOSX 10.9 with clang-3.5
+ Linux for Intel Xeon Phi with Intel-icc 15.0, 16.0
(gcc-4.9 compatibility layer)
o Geant4 10.3-beta-01 has been tested using CLHEP-2.3.3.0, required
for external installation of the CLHEP library.
Please refer to the Geant4 User Documentation:
http://cern.ch/geant4/support/userdocuments.shtml
for further information about using Geant4.
----------------------------------------------------------------------------
List of features and fixes included in this Beta release since 10.2.p02:
o Configuration:
-------------
+ CMake:
o Set CLHEP-2.3.3.0 as the required version for external CLHEP library.
o Updated version of data sets:
G4ENSDFSTATE-1.3.1, G4RadioactiveDecay-4.4.1 and G4TENDL-1.2.
Updated tags for 10.3-beta.
o Analysis:
--------
+ Fixed incompatibility with ROOT 5.x and 6.x formats.
+ Fixed compilation warnings on Windows-64.
+ Updated to g4tools version 1.27.4 (see History_tools).
o Event:
-----
+ Introducing G4MultiEventAction to allow multiple user actions
in the same job.
o Externals:
---------
+ Updated CLHEP module to version 2.3.3.0:
o Added use_atomic.h header to use std::atomic if compiler supports it.
o MixMaxRng: corrected initialisation and use of 'numberOfEngines'
when more than one threads calls the constructor.
Use simple loop for flatArray(); engine built-in function turns out
to be extremely slow and unefficient.
Use seed_spbox() for seeding MixMax with single seed.
Avoid hard-coded constant for initialisation of VECTOR_STATE_SIZE in
MixMaxRng, to directly use actual value of "N" from mixmax generator.
Corrected check for counter range in method getState().
Changed default number of N in mixmax generator to 17 (was 256).
Removed use of obsolete 'register' keyword.
Provided new skipping coefficients for the new value of N and for
the optional case N=8.
o HepRandom: revised settings of "defaults" to use a single
thread_local pointer.
o Geometry:
--------
+ magneticfield:
o Added new stepper classes G4BogackiShampine23 (BS23),
G4BogackiShampine45 (BS45) and DormandPrince745 (DP45),
implementing third order (BS23) and fifth order (BS45, DP45)
embedded RK tableaus.
o Added new stepper classes embedded RK method: DoLoMcPriRK34 (6-stage
3/4 RK, interpolation), DormandPrinceRK56 (9-stage 5/6 RK,
interpolation, FSAL-able) and DormandPrinceRK78 (13-stage 7/8 RK,
interpolation).
o Added TsitourasRK45 stepper.
o First version of FSAL classes: FSAL Integrator Driver (concrete,
stand-alone driver); FSAL Integrator Stepper (base class);
FBogackiShampine45 (FSAL-version of BogackiShampine45 stepper);
FDormandPrince745 (FSAL-version of FDormandPrince745 stepper).
o G4MagIntegratorStepper: added counter for calls to equation RHS,
with Get/Reset() methods.
o G4HelixMixedStepper: fixes and added new (5th order) stepper choices.
o Fix in G4CashKarp for the size of arrays holding intermediate values.
+ management:
o Added G4ScaleTransform class for Cartesian scaling transformations.
o Introduced G4BoundingEnvelope helper class to be used for the
calculation of the extent of a solid within the limits defined by
the G4VoxelLimits object.
o Properly use 3D transformations in G4ReflectedSolid. Removed unused
affine-transformation data and related methods; removed not used cached
inverse transformation. Simplified code.
Re-implemented CalculateExtent() to use new class G4BoundingEnvelope.
+ solids/Boolean:
o Added new class G4ScaledSolid providing ability to scale dimensions of
a shape in X, Y or Z.
o Boosted computation of GetPointOnSurface() for G4BooleanSolid;
reimplemented algorithm, using cached list of primitives.
o Added warning in G4BooleanSolid::GetPointOnSurface() in case of non
convergence after fixed number of trials.
+ solids/specific:
o Revised implementation for GetSurfaceArea() and GetCubicVolume() in
G4GenericTrap.
o Added 'const' qualifier for methods in G4ExtrudedSolid.
Added CheckPolygon() to remove collinear and coincident points from
polygons.
o Improved algorithms for computation of area in G4TriangularFacet and
G4QuadrangularFacet. Added post-const qualifier to GetArea().
o G4QuadrangularFacet: added exhaustive tests in constructor to catch
potential problems with a quadrangular facet: collinear vertices, non
planar surface, degenerate, concave or self intersecting quadrilateral.
o G4TriangularFacet: improved test in constructor to detect degenerate
(too small or too narrow) triangles.
o Global:
------
+ Added 'us' and 'ps' units symbols for microsend and picosecond
respectively to G4UnitsTable. Requiring new CLHEP library 2.3.3.0
for external CLHEP library.
+ Converted all units and constants from "static const" to "static
constexpr".
+ Introduced new utility class G4MTBarrier implementing a synchronization
point between threads.
+ Changed exception code in G4TWorkspacePool from G4TWorkspacePool
to "WorkspaceNN".
+ Changed date for release 10.3-beta.
o Materials:
---------
+ G4DensityEffectData, G4IonisParamMat: fixed density effect parameters
corrections for the case when NIST mean ionisation potential is not equal
to that parameter in Shternheimer table.
+ G4NistElementBuilder, G4NistMaterialBuilder: added access to the mean
ionisation potential; put liquid water in front of the list of NIST
materials (index=0), so pure materials can be easier accessed via atomic
number (Z > 0).
+ Added new class G4AtomicShells_EADL which extends shell data up to Z=120.
+ G4MaterialPropertiesTable: moved complex methods from inline to source.
+ G4UCNMicroRoughnessHelper: all methods declared as const.
+ G4NistManager, G4NistMaterialBuilder, G4NistElementBuilder: added new
const methods FindElement(), FindMaterial() and FindSimpleMaterial(),
for possible run-time use in MT mode.
+ G4Material: avoid deletion of G4MaterialPropertyTable.
+ G4Element: added method GetZ_asInt().
+ G4AtomicShells, G4SandiaTable: substituted asserts by G4Exception.
+ Code cleanup; adoption of nullptr, use of const/static.
o Particles:
---------
+ Updated properties of particles according to PDG-2015.
+ Introducing floating level base to G4Ions and related classes.
Migrated G4NuclideTable to the new floating level base scheme;
temporarily disable new scheme for G4IsotopeMagneticMomentTable.
+ Fixed G4IonTable::GetIon() with non-zero level.
+ Changed default life of ion and hyper nucleus from -1 to 0.
+ Fixed Coverity defects.
o Persistency:
-----------
+ GDML:
o Added ability to automatically export the names of sensitive detectors
as auxiliary information.
o Added ability to import and export scaled shapes. Extended GDML schema
by adding "scaledSolid" tag supporting scaled shapes.
o New GDML schema version 3.1.4 (files: gdml.xsd, gdml_solids.xsd).
o Physics Lists:
-------------
+ Builders:
o Disabled multiple-scattering at post-step in OrderingParameterTable.
+ Constructors:
o electromagnetic:
- Added SetScintillationTrackInfo() method to G4OpticalPhysics with
associated command in G4OpticalPhysicsMessenger.
Fixed memory leak in G4OpticalPhysicsMessenger.
- Partial fix for allowing calls to optical UI commands in "PreInit"
state. Partially addressing problem report #1832.
- G4EmStandardPhysics_option3, option4: added process of e+e- pair
creation by e+ and e-; enabled lateral displacement for muons and
hadrons.
- G4EmStandardPhysicsGS: set default range factor to 0.1.
- DNA and ModelActivator constructors: corrected model/process names;
adjusted high energy threshold of solvation (kill) in respect to the
energy range of the elastic model.
Added G4DNAElectronSolvation to all DNA constructors.
- Updated G4EmDNAPhysics_option5 and G4EmDNAChemistry.
- Updated G4EmDNAPhysics_option7 constructor.
- Updated G4EmModelActivator. Fixed activation of DNA models.
Use Uehara model with chemistry. Corrected elastic process name
for ions.
- Added flag and command for whether or not to call InvokeSD() method
of G4OpBoundaryProcess.
- Fixed G4Exception message in G4OpticalPhysicsMessenger.
- Fixed Coverity defects.
o decay:
- Added G4UnknownDecayPhysics constructor.
- G4RadioactiveDecayPhysics: enabled atomic de-excitation by the call
to EM parameters in ConstructProcess() method.
+ Lists:
o FTFP_BERT_TRV: Switch on the low-mass diffraction dissociation
used in G4HadronHElasticPhysics.
Use GS multiple scattering with the most precise "error-free"
stepping instead of the default (Opt0), for the purpose of studying
the impact of multiple scattering on hadronic showers.
o Moved code to be inline in G4RegisterPhysLists so that it gets
compiled as part of the G4PhysListRegistry compilation unit.
This ensures that all the physics lists declared to the registry are
referenced by it such that the linker will include them if libraries
are built statically.
o Electromagnetic Processes:
-------------------------
+ DNA:
o G4DNAScreenedRutherfordElasticModel: modified handling of energy
boundaries at init time. Deprecated SetKillBelowThreshold() method;
modified initialisation to prevent erasing/rewriting data.
o G4DNAChemistryManager: reversed logic in CreateWaterMolecule()
for interface with physics models.
o G4DNAChampionElasticModel: rearranged declaration of attributes.
Deprecated GetKillBelowThreshold() method; removed attributes;
modified Theta() method. Reset check ekin >= LowEnergyLimit().
Added missing computation of 'sigma'.
o G4DNAUeharaScreenedRutherfordElasticModel: added exception at init
time if model is used for anything other than e-.
Deprecated SetKillBelowThreshold() and GetKillBelowThreshold() methods,
use G4LowECapture or G4DNAElectronSolvatation; modified
SelectHighEnergyLimit() method; removed attributes.
Added possibility to extend applicability limit.
o G4DNAChampionElasticModel, G4DNAUeharaScreenedRutherfordElasticModel:
added pre-processor flag to control Verbose check mode; checking energy
boundaries in Initialise(), removed call to G4Electron; cleanup in
CrossSectionPerVolume by removing unnecessary checks and calls to map;
removed energy checks in SampleSecondaries().
o Updated fast computation of G4DNAScreenRutherfordElasticModel and
G4DNAUeharaScreenedRutherfordElasticModel.
o Renamed class G4DNAElectronSolvatation to G4DNAElectronSolvation.
o G4DNAOneStepThermalizationModel: using pre-processor flag to control
check mode verbosity; removed call to G4Electron at initialization.
o G4DNATransformElectronModel: using pre-processor flag to control check
mode verbosity; removed call to G4Electron at initialization; added
check for creation of solvated electron.
o G4MoleculeCounter: at initialization time, prevent creating entries
for unregistered molecule definition.
Removed check of c++ version for unique_ptr/auto_ptr selection.
o G4DNAMeltonAttachmentModel, G4DNASancheExcitationModel: removed usage
of internal energy boundaries.
o G4DNASancheExcitationModel: removed warning when SANCHE_VERBOSE is OFF.
o G4VUserChemistryList: added flag; some code cleanup.
o Added nrj extensibility to SR and USR models.
o Replaced usage of std::exp() by G4Exp().
+ High Energy:
o Use C++11 features explicit, override, delete, nullptr.
+ Muons:
o G4ePairProduction: new process e+e- pair production by e+-.
o Use C++11 features explicit, override, delete, nullptr.
+ Polarisation:
o G4PolarizedCompton, G4eplusPolarizedAnnihilation,
G4ePolarisedIonisation: in PostStepGPIL(), use cached values of
'currentInteractionLength'; fixes problem at boundaries where
updated value was incorrectly used.
o C++11 migrations: explicit, override, delete operators on function
declarations; remaining nullptr; write floats/doubles with '.'.
Use C++11 features nullptr and range-based for loop in
G4PolarizationManager.
+ Standard:
o G4UrbanMscModel: use new lateral displacement algorithm in opt-3
and opt-4 physics-lists. Modified randomisation of "true path length".
o G4UniversalFluctuation: fixed bug in Glandz part, Gaussian sampling.
The sampling is symmetrical around 'emean' now. Some code cleanup.
o G4SeltzerBergerModel: fixed data handling for Z > 92.
Addressing problem report #1870, about invalid read of data
at run time.
o G4eBremsstrahlungRelModel: change LPM constant such that it gives
suppression variable consistent with Migdal's one; fixes small issue in
LPM function computation.
o G4alphaIonisation, G4hIonisation, G4ionIonisation: do not define
StepFunction in constructor.
o Fixed shadowing compilation warnings in G4GoudsmitSaundersonMscModel
for 'ps'.
o G4ASTARStopping, G4PSTARStopping, G4NISTStoppingData, G4BraggModel:
data arrays become float, computation is left double.
o G4BraggModel, G4BraggIonModel, G4BetheBlochModel: use float for local
static data.
o Added c++11 keywords: explicit, final, override, delete; use nullptr
and use G4Element::GetZasInt().
+ Utils:
o G4VEmProcess, G4EmModelManager: added extra methods to access models.
o G4VEnegryLossProcess, G4EmCorrections: removed unused class members.
o Use C++11 keywords explicit, override, delete and nullptr.
o G4EmParameters, G4EmParametersMessenger: moved remaining parameters
from G4EmProcessOptions, which now becomes obsolete.
o G4EmCalculator, G4LossTableBuilder, G4LossTableManager, G4VMscModel,
G4VEmModel, G4VEmProcess, G4VEnergyLossProcess: use only G4EmParameters
class interfaces for initialisation.
o G4NuclearFormfactorType: added new enumerator.
o G4VMultipleScattering: moved position change to AlongStep.
Added method NumberOfModels().
o G4EmModelManager: reviewed and fixed model configuration for DNA
physics.
o G4EmParameters: allow to change parameters only from the master thread
and only at PreInit or Idle states.
o G4EmCorrections: make part of data static, make internal vectors static,
shared between threads.
o G4LossTableManager, G4VEmModel, G4VMultipleScattering: minor cleanup of
static const data.
o G4VEmModel, G4VAtomDeexcitation, G4EmCalculator: use new method
GetZasInt() from G4Element.
o G4VEmModel: fix in SelectRandomAtomNumber() method. Fixed destruction
in MT mode.
o G4VAtomDeexcitation: fixed possible data race at initialistion.
Extended printout at initialisation.
Completed migration to G4EmParameters interface.
o G4AtomicShell: make methods inlined.
o G4DummyModel: set default low-enegry limit to 0 (needed for DNA).
o G4VMultipleScattering, G4EmModelManager, G4LossTableManager: improved
debug printout.
o Fixed Coverity defects.
+ Xrays:
o Added new class G4ScintillationTrackInformation and use it in
G4Scintillation.
o Use C++11 features explicit, override, delete, nullptr.
o Generic Processes:
-----------------
- Biasing:
o Checked for unbounded while loop in G4ProcessPlacer.
+ Decay:
o Added better diagnostics in exception in G4Decay::DecayIt().
+ Optical:
o Added flag for InvokeSD() method call in G4OpBoundaryProcess.
o Hadronic Processes:
------------------
+ cross_sections
o Checked do-while loops for potential infinite loops.
+ management
o G4HadronicProcess: added flag for "integral" approach to take into
account cross-section change at a step. Do not create new materials
in GetElementCrossSection(), just try to find existing simple
G4Material ones instead. Adopt C++11 keyword nullptr.
o G4HadronicProcessStore: call initialisation of all
G4HadronicInteractions.
+ models/coherent_elastic
o G4LMsdGenerator: restored hadronic internal decay scheme allowing for
stable decay of the Roper resonances.
+ models/de_excitation
o G4EvaporationChannel, G4EvaporationProbability: fixed kinematics
and added discrete excitation energy for the residual fragment
if the excitation is below maximum known level energy.
o G4NuclearLevelData: added extra methods allowing to access the maximum
level energy per isotope in level DB. Added method FindLevel() allowing
to find discrete level for any consumer model.
o G4DeexPrecoParameters, G4NuclearLevelData: added new class for
parameters.
o G4EvaporationProbability: fixed bug in computation of probability
density function; use new method FindLevel() in G4NuclearLevelData when
sampling kinetic energy of emitted particles; changed interface for
sampling method. More detailed grid for probabilities.
Use same sampling algorithm as in the pre-compound model (do not
interpolate probability).
o G4EvaporationChannel: use new interface to the sampling of energy.
o Updated list of parameters; added full list of pre-coumpound parameters
and extra de-excitation ones. Allow to change parameters only from the
master thread and only at PreInit or Idle states.
o G4ExcitationHandler: allow addition of new evaporation or FermiBreakUp
models after the initialisation of the handler.
Removed G4Exception when MF model is called, instead increase default
limit on excitation energy per nucleon to 100GeV.
o G4ExcitationHandler, G4EvaporationChannel, G4EvaporationProbability,
G4NeutronRadCapture: use G4DeexPrecoParameters for initialisation;
initialisation is done once at the first event.
o G4NucLevel, G4LevelManager, G4LevelReader, G4PhotonEvaporation: changed
data structure for the transition data; instead of final energy, index
of final level, added multi-polarity of the transition.
This will allow for correlated gamma emission.
o G4EvaporationChannel, G4EvaporationProbability: improved initialisation
of run time members
o G4NuclearLevelData: improved FindLevel() method.
o G4VEvaporationChannel, G4LevelReader, G4ExcitationHandler,
G4LevelReader: removed commented lines, removed empty descructors.
o G4VEvaporation, G4Evaporation, G4VFermiBreakUp, G4FermiBreakUp,
G4PhotonEvaporation, G4ExcitationHandler: cleaned up initialisation;
added G4VFermiBreakUp pure virtual methods allowing to have alternative
models.
o G4Evaporation: allow change of the channel factory after initialisation.
o G4NeutronRadCapture: added initialisation of G4PhotonEvaporation.
o G4DeexPrecoParameters, Evaporation: use default channel factory.
o G4NuclearLevelData, G4LevelManager: added method to access low-edge
levels.
o G4KalbachCrossSection: more smooth inverse cross section for transitions
of high energy.
o Removed specific GEMCoulombBarrier classes per ion.
o G4PhotonEvaporation: rationalised initialisation in case radioactive
decay initialisation is incomplete, no effect on hadron inelastic
interactions.
o G4DeexPrecoParameters, G4NuclearLevelData, G4ExcitationHandler,
G4PhotonEvaporation: fixed initialisation in MT mode if physics-list
has only radioactive decay.
o Added c++11 keywords in many headers.
o Added warning in G4ExcitationHandler for super excited fragments;
improved warning printout for the high energy excitation.
o G4LevelManager, G4NucLevel: substituted asserts by G4Exception.
o Cleaned up code in CoulombBarrier and FermiFragment classes.
o Fixed Coverity defects.
+ models/management
o G4HadronicInteractionRegistry: added InitialiseModels() method.
G4HadronicInteraction: added InitialiseModel() method.
o G4VPreCompoundModel, G4VIntraNuclearTransportModel,
G4HadronicInteractionRegistry, G4HadronicInteraction:
added c++11 keywords, minor cleanup.
+ models/particle_hp
o Changed default option of GNUmake script; G4PHPDEBUG flag was
unintentionally activated.
+ models/parton_string
o Major revision of QGS model; implemented Reggeon cascading and
associated "Fermi motion". Implemented new algorithm of parton
kinematical properties determination and calculation of residual
nucleus properties. Adjusted G4GammaParticipants to the revised
model.
o Improved energy-momentum conservation in the QGS model, especially
in the case of diffraction interactions.
o G4QGSMSplitableHadron: fixed bug in the case of gamma projectile with
P_minus = 0.
o New tuning of FTF model parameters.
o Introduced various <Pt^2> for diffractive and non-diffractive
interactions in G4FTFModel, for NA49 and NA61/SHINE experimental data.
o Fine tuning in G4DiffractiveExcitation for the FTF model, to describe
NA49 experimental data, especially, <Pt> on Xf.
o Fix to address the baryon puzzle of NA49 data; <Pt> of baryons 1/2 at a
string fragmentation setting up at 435 MeV/c, and <Pt> for barion 3/2
production st 900 MeV/c.
o Fine tuning of G4LundStringFragmentation. <Pt> of mesons created at
quark fragmentation is decreased. <Pt> of baryons created at quark and
di-quark is increased. Corrected fragmentation functions of qq-> B(1/2)
and B(3/2). <Pt^2> is increased in 2 times.
Important for a description of NA49 experimental data on pp-interactions
at 158 GeV/c, especially for xF and Pt distributions.
o Introduced smearing of delta-isobar mass; delta-isobars are now treated
as a kinetic track; improved parameterisation of the quark exchange
process with excitation.
NA61/SHINE exp. data on Pi+, Pi-, K+, K- and protons in PP interactions
at 20, 31, 40, 80 and 158 GeV/c are now described. Now reproducing the
ratio of <Strane Q>/<Normal Q> in PP interactions. Not describing kink
in AA interactions.
o Improved di-quark fragmentation into baryons in Lund string
fragmentation. Improved description of NA49 and NA61/SHINE exp. data
on p+p->p+X. Improvements in method SetMinimalStringMass(), relevant
for Pbar P annihilation.
o Moved G4PomeronCrossSection class to 'qgsm' module.
+ models/pre_equilibrium
o G4PreCompoundModel: changed non-equilibrium condition to switch to
de-excitation; this allows to perform pre-equilibrium emission for
light nuclei.
o Migrated classes to use G4DeexPrecoParameters; removed obsolete class
G4PreCompoundParameters.
o G4GNASHTransitions: fixed problem in computation of probability.
o Use nullptr and define internal const as static const.
+ models/radioactive_decay
o G4RadioactiveDecay: renamed several methods and variables to more
closely represent their meaning.
In DoDecay(): added protection for zero pointer in case of missing
decay channel. Added new method AddDeexcitationSpectrumForBiasMode().
In BuildPhysicsTable(), removed initialization of atomic de-excitation
to avoid overwriting its initialization in EM.
Throw fatal exception if initialization not done somewhere.
Initialisation should be now done in physics-lists.
o Added production of gamma and electron spectra from de-excitation of
excited levels in the biased mode.
o Removed obsolete class G4RIsotopeTable.
+ models/util
o Fixed Coverity defect in G4Fragment.
o Run
---
+ Added handling of multiple actions to be instantiated and handled by
user code. Added G4MultiRunAction class to allow for multiple run actions
in a single job; added G4MultiPrimaryAction to allow for multiple
user-primary generation actions in a single job.
+ Renamed all MT related exceptions to have new code matching pattern.
+ Use scoped enum for G4MTRunManager::WorkerActionRequest and
added PROCESSUI action.
+ Added G4MTRunManager::GetNumberActiveThreads() method.
+ Use new class G4MTBarrier to implement synchronization points.
+ Introduced skeleton for new functionality to force workers to
execute UI commands w/o beamOn.
Introduced new UI command: "/run/workersProcessCmds" to trigger
execution of UI commands by workers even without a "/run/beamOn";
C++ API is G4MTRunManager::RequestWorkersProcessCommandsStack().
+ When workers are terminated they execute the UI commands issued
after the last /run/beamOn
+ Fixed potential memory leak in G4RunMessenger.
+ Disabled multi-scattering during post-step in G4PhysicsListHelper.
Fixed default process ordering parameters for G4UnknownDecay that
should not have AtRest.
+ G4AdjointSimManager: correction in the use of run actions to avoid
infinite loop during an adjoint simulation.
+ G4AdjointPrimaryGenerationAction: removed 3 times splitting of primary
adjoint gamma.
o Track & Tracking
----------------
+ Change G4VUserTrackInformation pointer in G4Track to mutable and setters
to be const-methods. This allows the user to set his/her
G4VUserTrackInformation pointer without casting G4Track const pointer.
+ New G4Multi{Stepping,Tracking}Action classes to allow for multiple
user-actions in the same job.
o Visualization:
-------------
+ management:
o MT mode: moved flush to end of visualization thread; this is so that
all trajectories are flushed from the thread before switching back to
master thread. Also fixing picking in MT mode.
o Introduced UI command /vis/viewer/interpolate.
o Added some calls to SetView for security.
o Inhibited use of BeginRun, BeginEvent, EndEvent and EndRun
when vis is not enabled, for example after /vis/disable.
In particular this prevents the start of the vis thread in
multi-threading mode.
o G4VisExecutive: registered drawing and filtering by touched volume.
o G4VisCommandsSceneAdd: removed incorrect check on existing models in
command "/vis/scene/add/volume". It was preventing parallel worlds being
added individually.
o Fixed mistake in introductory vis manager message.
o Improved error tracking: introduced error code to track error
conditions in G4VVisCommand; tracked error conditions for '/vis/open'
and '/vis/drawTree' in G4VisCommandsCompound.
o Improved '/vis/viewer/save'.
o Use SetWorldVolume() for OriginVolume drawing and filtering.
o Code cleanup in G4VisCommandsSceneHandler.
+ modeling:
o Introduced drawing and filtering by touched volume.
i.e., "/vis/modeling/trajectories/create/drawByTouchedVolume"
"/vis/filtering/trajectories/create/touchedVolumeFilter"
and their associated commands.
This selects only by a single string, the physical volume name.
If there are more than one physical volumes with that name then any
trajectory touching any one of those volumes will be selected.
If the user sets more than one physical volume name (by using the set
command more than once) the trajectory will be selected if it touches
either of those volumes and which of the attributes (e.g., colour)
associated with those volumes is chosen is arbitrary.
A more comprehensive facility would allow the user to select by
touched touchable, i.e., by the full path and copy numbers of the
physical volume - but that is a job for the future if a use case arises.
o Added specific arguments to LocateGlobalPointAndSetup() to handle
multi-threading mode in G4TrajectoryDrawByOriginVolume::Draw() and
G4TrajectoryOriginVolumeFilter::Evaluate().
o G4TrajectoryModelFactories: added
G4TrajectoryDrawByTouchedVolumeFactory.
o G4TrajectoryFilterFactories: added
G4TrajectoryTouchedVolumeFilterFactory.
o Added new files and classes: G4TrajectoryDrawByTouchedVolume and
G4TrajectoryTouchedVolumeFilter.
o G4ModelColourMap: sdded new access function GetBasicMap().
o G4ModelingParameters: added VisAttributesModifier set commands.
o Tidied uo G4TrajectoryDrawByOriginVolume.
o G4TrajectoryDrawByOriginVolume: tidied #include statements.
+ OpenGL:
o Xm: Ensure all items get flushed to screen, even when
'eventsDrawInterval' is set greater than 1.
o Fixed some inconsistencies in "/vis/ogl/set/eventsDrawInterval" and
"/vis/ogl/set/displayListLimit" commands.
o Fixed bug in picking output text.
o G4OpenGLXViewer: Added thread awareness regarding glXMakeCurrent.
o Fix for notifying multiple OGL X views.
o Fix in G4OpenGLViewer::getRealPrintFilename() so that files can
easily be referenced in alphanumeric order.
o CompareForKernelVisit(): tested for total visualization attributes
modifiers instead of just for size.
o Fixed typo for macro G4CONDITIONBROADCAST.
+ OpenInventor:
o G4OpenInventorXtExaminerViewer: designed new icons (XPM) for the
"Extended Pick Mode" and "Pick Reference Trajectory" buttons.
o Modified the mouse-over-trajectory output to make it more clear.
o Improved diagnostics for Bookmark file handling.
o New approach for determining distance along the reference path, using
element centers instead of bounding boxes. Solves problems with spurious
bounding box warnings and incorrect distance values.
o G4OpenInventorXtExtended & G4OpenInventorXtExtendedViewer: suppressed
some unwanted diagnostic output.
o Fixed setting of transparency type
(SoGLRenderAction::SORTED_OBJECT_ADD).
o Fixed dispatching.
o Fixed Coverity defects.
+ Tree:
o G4ASCIITreeSceneHandler: prevent multiple writing of the informational
header on the output file.
+ XXX:
o CompareForKernelVisit(): tested for total visualization attributes
modifiers instead of just for size.
o Data Sets:
---------
+ G4ENSDFSTATE-1.3.1:
o Added missing data for AC225, SI25, 193RN, TE140.
+ G4RadioactiveDecay-4.4.1:
o Added KM coefficients for e- capture and taken from DDEP data.
o Correction to Fe55
o Corrected 0 lifetime for Ac225.
+ G4TENDL-1.2:
o Compressed version of the library.
o Added information of source (ENDF/B-VII.1 or TENDL-2014) for each file.
o Fixed problem to Deuteron/Inelastic/F01/1_3_Hydrogen.z.
o Examples:
--------
+ Updated reference outputs, macros and scripts.
+ advanced/air_shower
o Replaced usage of std::log() by G4Log().
+ advanced/amsEcal
o PhysListEmStandard: use G4EmParameters.
+ advanced/composite_calorimeter (ccal-V10-02-00)
o Replaced usage of std::exp() by G4Exp().
+ advanced/hadrontherapy
o Added new modulator class.
+ advanced/microbeam (microbeam-V10-02-00)
o Replaced usage of std::exp() by G4Exp().
+ advanced/microelectronics
o Fixed multiple-scattering definition in PhysicsList.
+ advanced/nanobeam
o Replaced usage of std::exp() by G4Exp().
+ advanced/underground_physics
o DMXPhysicsList: migrated EM physics to G4EmParameters.
+ extended/electromagnetic/TestEm0
o PhysicsList: use G4EmParameters.
+ extended/electromagnetic/TestEm1
o PhysicsList::AddRadioactiveDecay(): set like threshold to
0.1*picosecond in G4NuclideTable. Updated radioactive.mac.
Use G4EmParameters.
o Added decayinfly.mac macro.
o Modified colors in vis.mac.
o PrimaryGeneratorMessenger: restore command SetDefaultKinematic.
Updated all macros.
+ extended/electromagnetic/TestEm2
o PhysicsList: use G4EmParameters.
+ extended/electromagnetic/TestEm3
o PhysicsList: use G4EmParameters.
+ extended/electromagnetic/TestEm5
o EventAction: corrected filling of histogram 1.
o PhysicsList: use G4EmParameters.
o Removed geant3 sub-directory.
+ extended/electromagnetic/TestEm7
o Fix in c2_functions.
o Fixed compilation warnings on gcc-6.1.
+ extended/electromagnetic/TestEm8
o TestParameters, Run, PrimaryGeneratorAction, PrimaryGeneratorMessenger:
reorganized printout of beam particle; use energy smearing as an
addition of noise, so smeared noise should be added in units of energy.
+ extended/electromagnetic/TestEm10
o Major example revision & clean-up: simplified class names; refactored
detector construction & its messenger classes; revised scoring, removed
unused code and replaced histograms with standard calls to g4tools;
added StackingAction which accounts XTR gamma, all gamma and all e-
spectra; changed to a modular physics list and introduced
TransitionRadiationPhysics builder; clean-up of the primary generator
action; added test_suite.sh script.
+ extended/electromagnetic/TestEm11
o PhysicsList: added radioactiveDecay process; use G4EmParameters.
o Removed G4DiscreteScattering.
o Run: compute energy deposit per absorber
o PrimaryGenerator: restored SetDefaultKinematic().
o Updated macro radioactive.mac.
+ extended/electromagnetic/TestEm12
o PhysicsList, PhysListEmStandard: use G4EmParameters.
+ extended/electromagnetic/TestEm13, TestEm14
o PhysicsList: use G4EmParameters.
o Removed geant3 sub-directory.
+ extended/electromagnetic/TestEm15
o PhysicsList: use G4EmParameters.
+ extended/electromagnetic/TestEm18
o PhysListEmStandard: use G4EmParameters.
+ extended/hadronic/Hadr06
o Added NE213 material.
o PhysicsList: added G4StoppingPhysics.
o EmStandardPhysics: use G4EmParameters.
+ extended/hadronic/Hadr07
o New example to survey energy deposition and particle's flux from
an hadronic cascade.
+ extended/medical/dna/clustering
o Added GNUmakefile and reference output.
+ extended/medical/dna/dnaphysics
o Added flagProcess value for electron capture.
o DetectorConstructor, DetectorMessenger: ensure that geometry is created
once; material may be changed for next runs.
o SteppingAction: added label for solvation.
o Added 2nd argument in SteppingAction for number of threads.
o /run/initialise moved from code to macro.
o DNA physics configuration moved from code to macro.
+ extended/medical/dna/microdosimetry
o Increase cuts of G4ElectronCapture to 7.4 eV to remain compatible
with the energy validity of the Champion's elastic model.
+ extended/medical/dna/range
+ extended/medical/dna/wvalue
o Changed physics constructor selection.
+ medical/electronScattering
o PhysListEmStandard: use G4EmParameters.
o PhysListEmStandard_WVI, SS: added SetMscThetaLimit().
+ extended/parallel/TBB
o Ported code following last updates to 'run' category.
+ extended/parallel/ThreadsafeScorers
o Corrected GNUmakefile.
+ extended/persistency/persistency/gdml/G01
o Added test case for scaled solids: scaledTube.gdml.
o Added missing gdml files to CMake installation script.
+ extended/polarisation/Pol01
o PhysicsList: move to use current method of ConstructParticle().
+ extended/radioactivedecay/rdecay01
o Initialize atomic de-excitation in PhysicsList::ConstructProcess()
instead of in G4RadioactiveDecay to avoid overwriting initializtion
in EM.
o Set world size to 2 cm and gun direction along X.
o Added survey of each nuclide in a given time window.
Print-out activity of each nuclide.
+ extended/radioactivedecay/rdecay02
o PhysicsList: register hadronic processes.
+ extended/runAndEvent/RE01
o Removed obsolete use of readout geometry.
o Added protection against a track sticking on the calorimeter inner
boundary.
o Use G4UnknownDecayPhysics constructor and removed ad-hoc physics-list.
+ extended/runAndEvent/RE04
o Added command "/vis/drawVolume worlds" to vis.mac.
----------------------------------------------------------------------------
ReleaseNotes/Patch4.10.1-3.txt
+201
View File
@@ -0,0 +1,201 @@
Geant4 10.1 - patch-03 Release Notes
------------------------------------
5 February 2016
List of fixes included in this public patch since the public release 10.1.p02:
o Configuration:
-------------
+ CMake:
o Updated Apple Clang detection to work on Xcode-7.
o Always use libc++ with Apple Clang so that std-supporting library is
used for C++11 mode on older Mac systems where libstdc++ is the default.
o Updated std-flag setting rules.
+ GNUMake:
o Updated X11 settings in MacOS platform.
o Added settings for C++ standards in platform specific scripts.
o Analysis:
--------
+ Fixed G4Analysis::Tokenize() which was failing when processsing a string
containing opening double quota without a closing one.
+ Fixed handling ntuples created in MT mode on master thread
This fixes the problem reported in hypernews (#459) and the problem
of deleting non-empty Root file containing only ntuples.
o Geometry:
--------
+ divisions:
o Fix for use of rotation matrix in G4VDivisionParameterisation in
MT-mode; use thread-local pointer for rotation matrix as member of the
class G4VDivisionParameterisation and set it to suto-delete.
Removed unnecessary thread-local verbose flag.
Addressing problem reports #1743 and #1758.
o Set as unsupported divisions of polycone and polyhedra in MT-mode.
+ magneticfield:
o Added check in G4ChordFinder against number of iterations in
FindNextChord(), to identify potential cases of lack of convergence.
o Better diagnostics text in case of errors in G4FieldTrack and
G4UniformMagField.
o Added explanation comment in G4MagHelicalSteper.
+ navigation:
o G4MultiLevelLocator::EstimateIntersectionPoint(): fix for logic error
in keeping consistent candidate intersection. Copes with case that a
candidate intersection is not found in the next segment (typically in
the 'second half' part).
Ensure that splitting of step (new level/depth) advances the required
length, used when adding a new level/depth.
Revised condition of 'while' loop which ensures that a new candidate
intersection exists. Corrected the resetting of 'first_section' flag.
Replaced fatal error for exceeding max steps with warning and signal
'looping' for too many steps; G4PropagatorInField detects this
condition, and identifies it as a 'Looping Particle' which is then
flagged; the calling class (Transportation, PathFinder) is expected
to stop the tracking of this particle after a few of these steps.
Addressing issue seen by ALICE (non-finishing of step due to poor
advances).
o G4VIntersectionLocator: added new logging methods & revised interfaces.
o G4NavigationLogger: added method to report issue with normal that
is the result of applying a rotation.
o G4NormalNavigation, G4VoxelNavigation: added check of rotated exit
normal.
o Relaxed condition for zero or almost-zero steps in G4ReplicaNavigation
and G4Navigator, to allow for faster progression in case of stuck tracks
in 3D scoring meshes. Addressing old problem report #1432.
o Fix in G4ReplicaNavigation::ComputeStep() for correct setting of copyNo
for particle entering. Fixes issue of negative copyNo observed in nested
replica setups. Addressing problem report #1634.
o Moved check for overlaps in G4Navigator::ComputeStep() within G4VERBOSE.
Addressing problem report #1811.
+ solids/Boolean:
o Checked for potential never-ending loops. Replaced while-do loop with
bounded for-loop in G4IntersectionSolid::DistanceToIn(p,v).
Also addressing problem report #1821.
+ solids/CSG:
o Introduced check in G4Sphere::DistanceToIn(p,v) for concave Theta and
point located on the origin. Also make proper use of radial tolerance in
DistanceToOut(p,v).
o Fixed Coverity defect in G4CutTubs::GetMaxMinZ().
+ solids/specific:
o Added specialised implementation for GetCubicVolume() in
G4TessellatedSolid. Added debug printouts withing G4SPECSDEBUG for
DistanceToIn/Out(p,v) to identify potential wrong responses.
o Removed redundant condition in G4Polyhedra and G4polycone
SetOriginalParameters() method.
o Minor code cleanup in G4SurfaceVoxelizer.
o Checked for potential never-ending loops.
+ volumes:
o Enabled paramerisation by solid type in G4GeometryWorkspace in MT-mode.
Assumes solids being parameterised are declared thread-local in the
user's parameterisation class and allocated just once.
o Global:
------
+ Tune radial tolerance to same value as for Cartesian tolerance in
G4GeometryTolerance.
+ Corrected conditions for proper treatment of gcc-5 in tls.hh and
G4Threading.hh.
+ Removed unused and buggy operator=(const G4String&) in G4SubString.
Added string length check in G4String::operator==().
Added missing 'inline' specification to functions in G4String.
+ Force use of G4MT wrappers for Intel-icc compiler for random numbers.
+ Updated date and version for 10.1.p03.
o Materials:
---------
+ G4NistElementBuilder: fixed typo in printout.
Addressing problem report #1786.
o Parameterisations:
-----------------
+ gflash: fix for Zeff wrongly computed as ( W1*Z2 ) + ( W2*Z1 ), with Z1
and Z2 swapped. Addressing problem report #1719.
o Particles:
---------
+ Fixed wrong argument order CreateIon(A,Z,LL,0.0) in G4IonTable to
CreateIon(Z,A,LL,0.0).
+ Fixed setting of mass in decay parent particle.
Addressing problem report #1820.
o Persistency:
-----------
+ GDML:
o Use relaxed precision constant for matrix to angle formula evaluation
in G4GDMLWriteDefine, to allow for proper treatment of singularities.
Fixes issues of misplaced volumes in exported geometries of complex
detectors.
o Rectify rotation matrices in reading and writing GDML files to reduce
spurious roundoff errors in stored values.
o Physics Lists
-------------
+ constructors/limiters:
o G4ParallelWorldPhysics: fixed process order index for
G4ParallelWorldProcess to 9900 to make sure it is registered
prior to G4OpBoundaryProcess.
o Processes - Electromagnetic:
---------------------------
+ xrays:
o G4VTransitionRadiation, G4VXTRenergyLoss: fixed process sub-type.
o Processes - Generic:
-------------------
+ scoring:
o G4ParallelWorldProcess: fixed processType from 'fParameterized'
to 'fParallel' and set processSubType to 491.
o Fixed Coverity defects.
o Processes - Hadronic:
--------------------
+ cross_sections:
o G4NeutronInelasticXS, G4NeutronCaptureXS: fixed isotope cross-sections
including retrieve, data management, isotope x-section and isotope
selection. Fixed Coverity defects.
+ models/cascade:
o G4CascadeCoalescence: added explicit inclusion of
G4InuclElementaryParticle header to address compiler error from Cray
(as reported in Configuration Hypernews Forum #1764).
+ models/coherent_elastic:
o G4DiffuseElastic: bug-fix in BuildAngleTable() to improve the angle
bin accuracy at high energies.
+ models/im_r_matrix:
o Fixed compiler warning in clang-3.6.
o Run:
---
+ G4PhysicsListHelper: add entry for G4ParallelWorldProcess.
+ G4RunManager, G4WorkerRunManager: fixed process order index of
G4ParallelWorldProcess to 9900 to make sure it is registered
prior to G4OpBoundaryProcess.
o Tracking:
--------
+ G4SteppingManager: fix to properly set 'OriginTouchableHandle' for the
primary track. Addressing problem report #1773.
o Examples:
--------
+ Updated reference outputs.
+ basic/B4:
o Restored the third parameter in G4UIsession construction, which was
lost when moving G4UIExecutive at the beginning of main().
Addressing problem report #1802.
----------------------------------------------------------------------------
Technical Notes
---------------
o This patch should be applied on top of release 10.1, 10.1.p01 or 10.1.p02.
o Technical notes distributed for release 10.1 are also applicable and
valid for this patch.
The code and rebuilt binary libraries for release 10.1.p03 are available
through our "Download" Web page.
Please refer to the Geant4 User Documentation for further information about
using Geant4.
ReleaseNotes/Patch4.10.2-1.txt
+251
View File
@@ -0,0 +1,251 @@
Geant4 10.2 - patch-01 Release Notes
------------------------------------
26 February 2016
List of fixes included in this public patch since the public release 10.2:
o Configuration:
-------------
+ CMake:
o Fixed configuration error occurring when using CLHEP-2.3.1.1 external
installation. Addressing problem report #1805.
o Updated data-sets to RadioactiveDecay-4.3.1 and G4ENSDFSTATE-1.2.1.
+ GNUMake:
o Added settings for C++ standards in platform specific scripts.
o Analysis:
--------
+ Fixed definition of /analysis/ntuple command directory which was created
implicitly with the commands and so defined without a guideline and not
deleted.
o Externals
---------
+ zlib:
o Fixed compilation warnings on clang-3.7 compiler.
o Geometry:
--------
+ magnetic-field:
o Fix for spin tracking of particles with negative magnetic moment.
Addressing problem report #1831.
+ navigation:
o Relaxed condition for zero or almost-zero steps in G4ReplicaNavigation
and G4Navigator, to allow for faster progression in case of stuck tracks
in 3D scoring meshes. Addressing old problem report #1432.
o Fix in G4ReplicaNavigation::ComputeStep() for correct setting of copyNo
for particle entering. Fixes issue of negative copyNo observed in nested
replica setups. Addressing problem report #1634.
o Moved check for overlaps in G4Navigator::ComputeStep() within G4VERBOSE.
Addressing problem report #1811.
+ solids/Boolean:
o Minor optimisation in Inside(p) for G4UnionSolid and G4SubtractionSolid.
+ solids/specific:
o Fix in G4GenericTrap for computation of extent.
Addressing problem report #1809.
Disabled debug warning in SurfaceNormal().
o Correction in G4ExtrudedSolid::AddGeneralPolygonFacets() for precision
treatment in construction of tessels to avoid cases of collinear points
and consequent degenerating triangles.
Addressing problem report #1777.
o Use pre-computed kCarToleranceHalf constant in G4ExtrudedSolid, now
protected in G4TessellatedSolid.
o Global:
------
+ Corrected conditions for proper treatment of gcc-5 in tls.hh and
G4Threading.hh.
+ Removed unused and buggy operator=(const G4String&) in G4SubString.
+ Updated date and version for 10.2.p01.
o Graphics Representations:
------------------------
+ Use "" instead of <> for internal header inclusion in
HepPolyhedronProcessor.h.
o Materials:
---------
+ Fixed rare data-race in G4MaterialPropertiesTable for optical physics.
+ G4NistMaterialBuilder: added lock to build of new NIST material; needed,
as some physics processes may create a new NIST material from worker
thread.
+ G4NistElementBuilder: fixed typo in printout.
Addressing problem report #1786.
o Particles:
---------
+ Fixed wrong argument order CreateIon(A,Z,LL,0.0) in G4IonTable to
CreateIon(Z,A,LL,0.0).
+ Fixed setting of mass in decay parent particle.
Addressing problem report #1820.
+ Fixed G4DecayTable::SelectADecayChannel() for decay channels all
kinematically forbidden.
+ In G4IonTable removed static const tolerance and replaced with call to
G4NuclideTable; this recovers missed levels when they are closer together
than 2 keV. Reverted default value of level tolerance in G4NuclideTable
to 1 eV; turn out rounding of excitation energy to avoid potential
conservation problems.
o Persistency:
-----------
+ GDML:
o Corrected broken URLs and version number to schema.
o Physics Lists
-------------
+ builders:
o G4NeutronHPBuilder: corrected name from "NeutronHPInelastic" to
"ParticleHPInelastic".
+ constructors/limiters:
o G4ParallelWorldPhysics: fixed process order index for
G4ParallelWorldProcess to 9900 to make sure it is registered
prior to G4OpBoundaryProcess.
+ constructors/electromagnetic:
o G4EmStandardPhysicsSS: do not compute cross-section tables
if experimental Mott cross-section is enabled (this provides a
significant increase in CPU, but only in this conditions Mott
corrections are valid).
+ constructors/hadron_elastic:
o Added missing hadron elastic for He3 in G4HadronElasticPhysics,
using Glauber-Gribov elastic cross-section (Gheisha does not have
hadronic - elastic and inelastic - cross-sections for He3).
o Processes - Electromagnetic:
---------------------------
+ dna:
o In ionisation models SampleSecondaries(), added protection on energy
in case of interpolation method change.
+ lowenergy:
o Fixed inclusion guard in G4LivermorePolarizedPhotoElecticModel.hh.
Addressing problem report #1816.
o Code cleanup and fixed Coverity defects.
+ standard:
o Fixed cases of floating-point exception in G4ScreeningMottCrossSection.
Some code cleanup and performance improvements.
o Fixed sampling of scattering angle in G4eSingleCoulombScatteringModel.
Addressing problem report #1823.
o G4GoudsmitSaundersonMscModel G4GoudsmitSaundersonTable: updated step
limitations. Fixed Coverity defects and improved comments.
+ utils:
o G4EmSaturation: corrected method name from VisibleEnergyDeposition() to
VisibleEnergyDepositionAtAStep(), in order to allow overloading of
virtual method. Addressing problem report #1807.
o G4EmCalculator: fixed computation of dEdx and range for He3 and alpha.
Addressing problem report #1808.
o G4LossTableManager: avoid double addition of He3 and alpha to the loss
table map (should not have any effect on results).
+ xrays:
o G4VTransitionRadiation, G4VXTRenergyLoss: fixed process sub-type.
o G4Scintillation: use new method introduced in G4EmSaturation
VisibleEnergyDepositionAtAStep() in order to fix overloading problem.
o Processes - Generic:
-------------------
+ biasing:
o Code cleanup and Coverity fixes in 'generic' and 'management' modules.
+ scoring:
o G4ParallelWorldProcess: fixed processType from 'fParameterized'
to 'fParallel' and set processSubType to 491.
+ optical:
o Fix to DielectricLUT concerning the sequence of rotations.
Addressing problem report #1826.
o Processes - Hadronic:
--------------------
+ models/abla:
o Fixed bug in de-excitation of boosted nucleon remnants.
+ models/de_excitation:
o G4GammaTransition: added protection against rare problem in sampling
of internal conversion when transition energy is lower than atomic
level energy: instead of stopping the cascade, set bound level energy
to zero and produce final e-; this allows to complete radioactive-decay
cascade. Addressing problem report #1806.
o G4PhotonEvaporation, G4LevelManager, G4LevelReader, G4NuclearLevelData:
fixed remaining cases of production of excited stated.
Addressing problem report #1806.
o G4PhotonEvaporation, G4LevelReader, G4NucLevel: fixed treatment of
X-levels (meta-stable nuclear levels which have forbidden gamma
transition but opened beta-transitions); this fix excludes creation of
very low-energy fake gamma and provides a correct nuclear decay
sequence.
o G4PhotonEvaporation: set limit on emitted gamma energy 0.1 keV (was
10 keV).
+ models/lend:
o Fixed problem in prodcution of residual neutrons.
o Corrected use of isnan() to std::isnan() for compilers different than
Windows/VC++.
o Fixed Coverity defects.
+ models/particle_hp:
o Fixes for problem on materials with hydrogen or helium.
o Fixed problems in deletion of cross-section data set in MT mode
and initialisation of cached data.
o Fixed Coverity defects.
+ models/qmd:
o Fixed bug in handling forbidden decay in G4QMDCollision.
+ models/radioactive_decay:
o G4BetaMinusDecay, G4RadioactiveDecay: added protection for case
when end point energy is 0 in beta- decay.
Addressing problem report #1766.
+ models/util:
o G4KineticTrack: replaced G4cerr with a 'JustWarning' G4Exception.
+ util:
o G4Nucleus: increased the max number of loops allowed in the method
GetBiasedThermalNucleus().
o Run:
---
+ G4PhysicsListHelper: added entry for G4ParallelWorldProcess.
+ G4RunManager, G4WorkerRunManager: fixed process order index of
G4ParallelWorldProcess to 9900 to make sure it is registered
prior to G4OpBoundaryProcess.
o Tracking:
--------
+ G4SteppingManager: fix to properly set 'OriginTouchableHandle' for the
primary track. Addressing problem report #1773.
o Data sets:
---------
+ Patched data set for nuclides properties G4ENSDFSTATE-1.2.1:
o Added missing data for RA228, SI24, SI25 and 193RN.
+ Patched data set for radioactive-decay, G4RadioactiveDecay-4.3.1:
o Replace NaN values for intensity by 1 in z71.a157(Lu157),
z55.a136(Cs136) and z59.a129(Pr129).
o Examples:
--------
+ Updated reference outputs.
+ basic:
o Added G4ENSDFSTATEDATA in the lists of mandatory data sets; updated
README files. Addressing problem report #1810.
o B4: restored the third parameter in G4UIsession construction, which
was lost when moving G4UIExecutive at the beginning of main().
Addressing problem report #1802.
+ extended/biasing/B03:
o B03PhysicsList: corrected process order index of G4ParallelWorldProcess
to 9900 to make sure it is registered prior to G4OpBoundaryProcess.
+ extended/electromagnetic:
o Fixed macros to allow to work both with Root-5 and 6.
+ extended/radioactivedecay/rdecay01:
o PhysicsList: SetLevelTolerance to 1*eV, there appears to already be a
level Ni60[1332.510] stored somewhere instead of Ni60[1332.514].
Setting the tolerance to 1 eV gets the right level.
o Improved printout in Run::EndOfRun().
o Updated references in it1.mac and it2.mac.
----------------------------------------------------------------------------
Technical Notes
---------------
o This patch should be applied on top of release 10.2.
o Technical notes distributed for release 10.2 are also applicable and
valid for this patch.
The code and rebuilt binary libraries for release 10.2.p01 are available
through our "Download" Web page.
Please refer to the Geant4 User Documentation for further information about
using Geant4.
ReleaseNotes/Patch4.10.2-2.txt
+258
View File
@@ -0,0 +1,258 @@
Geant4 10.2 - patch-02 Release Notes
------------------------------------
17 June 2016
List of fixes included in this public patch since the public release 10.2.p01:
o Configuration:
-------------
+ CMake:
o Enhanced advanced option for use of external USolids/VecGeom library;
it is now possible to select individual shapes for partial use of the
set of primitives provided by the library, by specifying through the
GEANT4_USE_USOLIDS the list of primitives to consider, i.e. using:
-DGEANT4_USE_USOLIDS="box;tubs" will replace box and tubs shapes with
USolids/VecGeom ones; -DGEANT4_USE_USOLIDS="all" or -DGEANT4_USE_USOLIDS
will replace with all shapes enabled in the library.
o Fix for treatment of FindCLHEP module for the case of external CLHEP
installation. Addressing problem report #1855.
o Adding libXt to X11_LIBRARIES on all Unix platforms when using
Motif/Inventor. Addressing problem report #1828.
Added missing Xmu to X11_LIBRARIES when using Motif/OpenInventor on
MacOSX.
o Corrected Modules/genwindef/LibSymbolInfo.cpp for use with Windows
VC++ 64 bits compilation.
o Updated data-sets to RadioactiveDecay-4.3.2 and G4ENSDFSTATE-1.2.3.
+ GNUMake:
o Added flags for partial use of USolids/VecGeom types.
o Added Linux-clang.gmk script for use of clang compiler on Linux system.
o Analysis:
--------
+ Fixed compilation warnings from gcc-6.1.
o Digits-Hits:
-----------
+ Fixed typo in printing on cout in G4SDManager::AddNewDetector().
+ Fixed shadowing compilation warnings in G4VScoringMesh for 'ps'.
+ Code cleanup and introduced some C++11 features.
+ Checked do-while loops for potential infinite loops.
o Event:
-----
+ Fixed typo in printout. Addressing problem report #1860.
+ Fixed shadowing compilation warnings for 'ps'.
+ Code cleanup and introduced some C++11 features.
o Externals
---------
+ zlib:
o Fix in gzguts.h for Windows with Visual Studio 2015 in 64 bit
compilation.
o Geometry:
--------
+ magnetic-field:
o Corrected implementation of operator=() and copy-ctor for G4Field,
G4ElectroMagneticField, G4ElectricField and G4CachedMagneticField.
Fixed Clone() signatures in derived Field classes.
+ management:
o Throw exception in call to CreatePolyhedron() for base class G4USolid.
Method is now implemented in the concrete wrappers. Code cleanup.
o In G4USolid, directly call GetEntityType() from wrapper classes for
CreatePolyhedron().
o Added flag to USolids wrappers for partial use of USolids/VecGeom types.
o G4AffineTransform: added inline keywords and fixed formatting.
+ navigation:
o Avoid static constant definition of surface tolerance in
G4VoxelNavigation in method LocateNextVoxel() to clear spurious
warnings for data-race conflicts from Valgrind/DRD tool; defined
as class data member. Cleared up use of static constants also in
other navigation classes.
o Improved diagnostics in G4Navigator. Moved some debug checks to
G4VERBOSE.
o Fixed few compilation warnings on gcc-6.1.
+ solids/Boolean:
o Moved accessors and modifiers in UMultiUnion wrapper to be not inlined.
Required modification in order to allow correct dispatching to VecGeom
types.
+ solids/CSG:
o Added overloaded CreatePolyhedron() method to USolids wrappers.
It is now possible to visualize supported VecGeom shapes.
Added flag for partial use of USolids/VecGeom types.
Moved accessors and modifiers to be not inlined. Required modification
in order to allow correct dispatching to VecGeom types.
Added overloaded GetEntityType() method.
o G4Orb: fixed typo in G4Exception message.
+ solids/specific:
o Fix in G4IntersectingCone for smaller precision constant in
LineHitsCone1() and LineHitsCone2(). Fixes issue observed in
G4GenericPolycone. Addressing problem report #1794.
o Added overloaded GetEntityType() method to USolids wrappers.
Moved accessors and modifiers to be not inlined. Required modification
in order to allow correct dispatching to VecGeom types.
Added overloaded CreatePolyhedron() method. It is now possible to
visualize supported VecGeom shapes.
Added flag for partial use of USolids/VecGeom types.
o Global:
------
+ Define G4GeometryTolerance singleton instance pointer as tread-local
to get rid of spurious race conflicts at initialisation.
+ Added check for support of atomics for clang compiler in HEPRandom
wrappers.
+ Improved algorithm for G4RandomDirection.
+ Turn warnings to fatal exceptions in G4UnitsTable for non existing
units or categories in the table. Addressing problem report #1843.
+ Updated date and version for 10.2.p02.
o Interfaces:
----------
+ Fix for Qt driver for position of 'OK' button.
+ Fixed compilation warnings on MacOS for conversion to char* in G4UIXm.
+ Fixed compilation warnings in G4UIWin32 for Windows.
o Particles:
---------
+ Fixed MT data-race in G4VDecayTable. Addressing problem report #1834.
o Persistency:
-----------
+ GDML:
o G4GDMLWrite: increased max_chars in the transcode() calls.
Addressing problem report #1847.
o Added checks for correctness of imported units.
Addressing problem report #1842.
o Parameterisations:
-----------------
+ gflash:
o Added missing inclusion of header in G4VGFlashSensitiveDetector.
Addressing problem report #1867.
o Physics Lists
-------------
+ constructors/hadron_elastic:
o Made G4HadronElasticPhysicsPHP constructor thread-safe.
+ constructors/hadron_inelastic:
o G4HadronPhysicsFTFP_BERT_ATL: new constructor for ATLAS, similar to
G4HadronPhysicsFTFP_BERT but with the transition between BERT and FTFP
in the energy region [9, 12] GeV.
o G4HadronPhysicsQGSP_BIC_AllHP: treat only proton with ParticleHP in
this class, and move the others (deuteron, triton, 3He, alpha) to
the class G4IonPhysicsPHP.
+ constructors/ions:
o Added G4IonPhysicsPHP constructor class.
+ lists:
o New experimental physics-list FTFP_BERT_ATL, similar to FTFP_BERT but
with the transition between BERT and FTFP in the energy region [9, 12]
GeV. Meant to reduce jet energy scale difference observed in ATLAS.
o QGSP_BIC_AllHP : use new, specialized ion constructor, G4IonPhysicsPHP.
o Processes - Electromagnetic:
---------------------------
+ dna:
o Fixed compilation warnings from gcc-6.1.
+ lowenergy:
o Updated G4LowEPComptonModel; addressing problem report #1676.
o Replaced usage of std::exp() by G4Exp().
o Fixed compilation warnings in gcc-6.1. Code cleanup.
+ standard:
o G4eSingleCoulombScatteringModel, G4ScreeningMottCrossSection: fixed
initialisation of classes allowing to use precomputed cross-sections.
Addressing problem report #1823.
+ utils:
o G4EmCalculator: added protection against zero cut value; use lowest
electron energy. Addressing problem report #1845.
o Processes - Hadronic:
--------------------
+ models/de_excitation:
o G4PhotonEvaporation: fixed final energy in the case of X-level.
o G4ExcitationHandler: substituted usage of iterators by simple loop;
added exception for the case of infinite loop of evaporation.
Addressing problem report #1837.
+ models/lend:
o Fixed problem on setting birth time of delayed neutrons.
+ models/particle_hp:
o Fixed problem in deletion of models in multi-thread mode.
o Corrected message about a behavior under the activation of fission
fragment production.
o Fixed bug in usage of G4Nucleus::GetBiassedThermalNucleus().
o Fix for calculation of residual A and Z, affecting results for
charged particle projectiles. Addressing problem report #1815.
o Bug fix in sampling of kinematic energy of scattered particle in
DiscreteTwoBody. Addressing problem report #1815.
o Fixed warning message from gcc-6.1.
o Fixed Coverity defects.
+ models/parton_string/diffraction:
o Fixed bug in G4FTFModel for nucleus-nucleus interactions
Also improvements in sampling of kinematical variables.
o Improved G4FTFAnnihilation; now creating Pi0, Eta and Eta_prime.
+ models/quasi_elastic:
o Fixed shadowing compilation warnings in G4QuasiElRatios for 'ps'.
+ models/radioactive_decay:
o Use new API for G4VDecayChannel class to avoid MT data-race.
Fixes issues of observed event non-reproducibility in MT mode.
+ models/util:
o G4Fragment: introduced "spin", added accessors and modifiers needed
by INCL++; fixes a memory leak discussed on HN PhysicsList Forum #930.
o Use modified G4VDecayChannel API to avoid data-race in MT.
o Fixed Coverity defects.
o Track:
-----
+ Use G4Log in G4VelocityTable::Value() method.
o Visualization:
-------------
+ Fixed compilation warnings from gcc-6.1.
+ gMocren:
o Fixed compilation warnings on win64.
+ OpenInventor:
o Corrected delete statement in sortBookmarksCB() for
G4OpenInventorXtExaminerViewer.
o Fixed compilation warnings on MacOS for conversion to char* in
G4OpenInventorXtExtendedViewer and G4OpenInventorXtViewer sources.
o Data sets:
---------
+ Patched data set for nuclides properties G4ENSDFSTATE-1.2.3:
o Corrected packaging for download. Contents unchanged.
Addressing problem reports #1849, #1854.
+ Patched data set for radioactive-decay, G4RadioactiveDecay-4.3.2:
o Corrected packaging for download. Contents unchanged.
Addressing problem reports #1849, #1854.
o Examples:
--------
+ Updated reference outputs.
+ advanced/human_phantom:
o Fixed compilation warnings from gcc-6.1.
+ basic/B3:
o Fixed initialization of the hits collections IDs in B3aEventAction.
Addressing problem report #1862.
+ basic/B5:
o Fixed wrong determination of the copy number in replicas.
+ extended/field:
o field03, field04: fixed compilation warnings from gcc-6.1.
+ extended/parameterisations/Par02:
o Added magnetic field and corrected the way to create a G4FieldTrack
object in Par02FastSimModelTracker.
----------------------------------------------------------------------------
Technical Notes
---------------
o This patch should be applied on top of release 10.2.
o Technical notes distributed for release 10.2 are also applicable and
valid for this patch.
The code and rebuilt binary libraries for release 10.2.p01 are available
through our "Download" Web page.
Please refer to the Geant4 User Documentation for further information about
using Geant4.