For MT compatibility, we need to get rid of event-dependent data held in
member variables in the intersector tools.  In some cases, we don't
really need to save this across calls, but it turns out that for
SolenoidalIntersector, we do need to save it.  So allow attaching
an arbitrary data block to TrackSurfaceIntersection in which we can
stash this information.

This block will be something deriving from IIntersectionCache; we hold
a unique_ptr to it.

John Derek Chapman authored 6 years ago and Atlas Nightlybuild committed 6 years ago

Considering all surfaces (not only material ones) in TimedExtrapolator.cxx

See merge request atlas/athena!20694

(cherry picked from commit b95e0dcc)

e6704edd Considering all surfaces (not only material ones) in TimedExtrapolator.cxx to...

Do not set default values in initialize in GeantFollowerMSHelper.cxx because they are already set in BeginEvent.

Do not set default values in initialize in GeantFollowerHelper.cxx because they are set to default already at BeginEvent.

The tool was caching event information in member variables.
Rework to remove this caching, to allow the interfaces to eventually
be made const.
Results are identical, and this is not expected to have any significant
effect on performance.

Also changed CRLF line endings to LF.

In C++11, implicit declarations of copy and assignment are deprecated if the
class has a user defined destructor or copy or assignment.  gcc9 now warns
about this by default.
Adjust to avoid the warning.

Add a simple unit test for StraightLineIntersector.

All units migrated except GeoModelKernelUnits::gram/g, which is different from Gaudi::Units::gram/g

Minor updates BTagging config

The intersection tools currently start with a TrackSurfaceIntersection object,
cache the information from it in member variables, step the path through
the detector, and then create a new TrackSurfaceIntersection object.

We want to make the intersector interfaces const, which means getting
rid of the information cached in tool member variables.  We can streamline
this by keeping the stepping state in a TrackSurfaceIntersection object
directly, rather than in discrete member variables.  This allows eliminating
some copies at the end.

Add (non-const) interfaces allowing retrieving non-const references
to the member data of TrackSurfaceIntersection.

In C++11, implicit declarations of copy and assignment are deprecated if the
class has a user defined destructor or copy or assignment.  gcc9 now warns
about this by default.
Adjust to avoid the warning.

In C++11, implicit declarations of copy and assignment are deprecated if the
class has a user defined destructor or copy or assignment.  gcc9 now warns
about this by default.
Adjust to avoid the warning.

Remove SiDetectorManager.h/SCT_DetectorManager.h if possible. Migrate SiDetectorManager.h to SCT_DetectorManager.h and/or PixelDetectorManager.h if possible.

Non-members should not start with m_.

This commit updates the TP converters in `TrkTruthTPCnv` to be able to
write as well as read persistent formats using `HepMcParticleLink_p2`.
These changes were pullled in from `TrkTruthTPCnv-01-00-07`.
It also pulls in changes from `TrkTruthAthenaPool-00-02-06` to write
the persistent formats using `HepMcParticleLink_p2` by default.
This is part of the `HepMcParticleLink` migration - see ATLASSIM-2430.

In addition unit tests for some of the classes which use
`HepMcParticleLinkCnv_p1` and `HepMcParticleLinkCnv_p2` are added.
Namely `TrackTruthCollectionCnv_p1`, `TrackTruthCollectionCnv_p2`,
`TruthTrajectoryCnv_p1`, `TruthTrajectoryCnv_p2`,
`PRD_MultiTruthCollectionCnv_p2` and `PRD_MultiTruthCollectionCnv_p3`.
Still need to solve a set up issue for the
`PRD_MultiTruthCollectionCnv_p1` unit test, so not adding that yet.

Remove reference to nonexistent package HighPtBTrackingAnalysis.

Remove MA27 and clean up private member variables

See merge request atlas/athena!11334

(cherry picked from commit 974faa2b26cb119cf525a41e2fa7432d0bdd0a8c)

1139d41a Remove MA27 and clean up private member variables
9a68c581 Remove all references to MA27
fce0467e Remove reference to MA27 from iPatFitter
51c40550 Remove final reference to MA27
449fc553 Update CMakeLists file
014fa0a3 Merge remote-tracking branch 'upstream/21.0' into Remove_MA27
a4936b8d Change solver so something more suitable for the alignment like problems
2ee38b8d code cleanup -> NULL to nullptr
df76f86b start removing cout/cerr
7b7a6af6 Remove the remaining cout/cerr calls

Files with identical content (same SHA1) but different file permissions
cause confusing differences shown in GitLab. The file permissions of
61 files were set to 644 (ATLINFR-2011).

Sofia Adorni Braccesi Chiassi authored 6 years ago and Edward Moyse committed 6 years ago

Changes in the ambiguity solver to improve tracking for high Pt Bs. Changes can be turned on by using two different flags.


Former-commit-id: 363b497d

Given the standalone library is based on Eigen instead of CLHEP, all GeoModel clients
in Athena code were also migrated to the Eigen-based interface of GeoModel. This affected
a large number of packages. Most significant changes have been introduced to subsystem
GeoModel packages.

The in-house GeoModelKernel code has been removed from the repository.

Almost all header files in PixelGeoModel have been moved to the src directory.
The next step should be turning this package into a component-only library.
Similar code reorganization can be applied to SCT_GeoModel too.

Subsystem readout geometry interfaces remain mostly unchanged, except for a minor
change in the InDet readout geometry interface (which was unavoidable).

Several non-GeoModel packages affected by this migration mostly required addition
of missing CLHEP includes.


Former-commit-id: 9cffebb1

clang warnings: Missing override keywords.


Former-commit-id: 4a7bdb2c708852e7a45d3f21ba110a7dcfc72f3a

Charles Leggett authored 6 years ago and Vakhtang Tsulaia committed 6 years ago

Former-commit-id: 645176ec

Adam Edward Barton authored 6 years ago and Adam Edward Barton committed 6 years ago

Former-commit-id: 58aa95a8

Adam Edward Barton authored 6 years ago and Adam Edward Barton committed 6 years ago

Former-commit-id: 99bf3212

Former-commit-id: 6765fdd1

Christos Anastopoulos authored 6 years ago and Vakhtang Tsulaia committed 6 years ago

Former-commit-id: 0a174dcf03eecadfb60b55e3b11773149be9676e

Former-commit-id: 6b28f423

Former-commit-id: 570ddfe2

Correct error in vertex covariance matrix in case of simultaneous use of "Pass near PV" and any other constraints


Former-commit-id: 9194827d760ca6502400ca12f783ec62cb9b0745

Former-commit-id: 344e19f9

Former-commit-id: 450c901c

Adam Edward Barton authored 6 years ago and Walter Lampl committed 6 years ago

Former-commit-id: d07cad01