athena merge requestshttps://gitlab.cern.ch/atlas/athena/-/merge_requests2022-08-05T17:34:25+02:00https://gitlab.cern.ch/atlas/athena/-/merge_requests/55646Add base xaod container2022-08-05T17:34:25+02:00Tomasz BoldAdd base xaod containerThis is another attempt for: !55448
As advised by @mnowak - splitting dictionaries fixes the deadlock issue.
This MR bring the definition of the xAOD::BaseContainer. That is a container that has no actual content. It allows though to be...This is another attempt for: !55448
As advised by @mnowak - splitting dictionaries fixes the deadlock issue.
This MR bring the definition of the xAOD::BaseContainer. That is a container that has no actual content. It allows though to be freely decorates. It turned out that in several areas people abuse other containers for exactly that use-case, see discussion: [https://groups.cern.ch/group/hn-atlas-offlineSWHelp/default.aspx?RootFolder=%2Fgroup%2Fhn%2Datlas%2DofflineSWHelp%2FLists%2FArchive%2FEMD%20class%20for%20an%20objet%20having%20only%20decorations&FolderCTID=0x012002002BD1887D3A9C0D489560C010AA1FE9F4&View={5121AF6B-9AFF-4225-8B52-9244D9184E7C}](https://groups.cern.ch/group/hn-atlas-offlineSWHelp/default.aspx?RootFolder=%2Fgroup%2Fhn%2Datlas%2DofflineSWHelp%2FLists%2FArchive%2FEMD%20class%20for%20an%20objet%20having%20only%20decorations&FolderCTID=0x012002002BD1887D3A9C0D489560C010AA1FE9F4&View=%7B5121AF6B-9AFF-4225-8B52-9244D9184E7C%7D)Tomasz BoldTomasz Boldhttps://gitlab.cern.ch/atlas/athena/-/merge_requests/55697Daily "22.0 to master" sweep: 05/08/222022-08-08T08:12:01+02:00Vakhtang TsulaiaDaily "22.0 to master" sweep: 05/08/22The only conflict was in two `version.txt` filesThe only conflict was in two `version.txt` fileshttps://gitlab.cern.ch/atlas/athena/-/merge_requests/55737Get TrigEventSelectionAlg working in AthAnalysis2022-08-10T21:51:59+02:00Dan GuestGet TrigEventSelectionAlg working in AthAnalysisThis is a "fix" for an issue I was getting running a few algorithms that used `CutFlowSvc` in `AthAnalysis`.
On the first event I get
```
CutFlowSvc ERROR Inconsistent weights and variation sizes...This is a "fix" for an issue I was getting running a few algorithms that used `CutFlowSvc` in `AthAnalysis`.
On the first event I get
```
CutFlowSvc ERROR Inconsistent weights and variation sizes 217 and 1
terminate called after throwing an instance of 'std::runtime_error'
what(): Inconsistent weights and variation sizes
```
I'm tagging @tadej, @khoo, @jburr, @frenner, @viruelas. I'll work on an MRE, for now the problem is demonstrated [in some analysis code](https://gitlab.cern.ch/dguest/hh4b-analysis/-/blob/e18dc5ef3de0feccf63600c76e64e2b6df1eb57f/HH4bAnalysis/share/VariableDumperConfig.py).
Roughly speaking, I set up `MainServicesConfig`, `SystematicsSvc`, `TrigDecisionTool`, `TrigEventSelectionAlg`, and call `makeEventSelectionAnalysisSequence`.
Does anyone have this working in `AthAnalysis`?Dan GuestDan Guesthttps://gitlab.cern.ch/atlas/athena/-/merge_requests/55667Fixes for TruthJet writing during Gen_tf.py jobs2022-08-11T15:18:53+02:00John Derek ChapmanFixes for TruthJet writing during Gen_tf.py jobsIt was observed that despite being created during generation jobs `xAOD::JetContainer` objects for Truth jets were not being written out to the EVNT file. After some debugging a number of issues were identified and are fixed in this merg...It was observed that despite being created during generation jobs `xAOD::JetContainer` objects for Truth jets were not being written out to the EVNT file. After some debugging a number of issues were identified and are fixed in this merge request.
Update syntax for adding Truth `xAOD::JetContainer` objects to `StreamEVGEN`.
Block loading dictionaries which are not present in `AthGeneration` or `AthSimulation` during `xAOD::JetContainer` writing.John Derek ChapmanJohn Derek Chapmanhttps://gitlab.cern.ch/atlas/athena/-/merge_requests/55989ATR-25904: fixed the RUN3_ROI_OVERFLOW_SHIFT and m_OverflowPerRoIMask to 122022-08-18T21:21:19+02:00Patrick Karl CzodrowskiATR-25904: fixed the RUN3_ROI_OVERFLOW_SHIFT and m_OverflowPerRoIMask to 12Fixing the bit shifts as described in ATR-25904Fixing the bit shifts as described in ATR-25904Patrick Karl CzodrowskiPatrick Karl Czodrowskihttps://gitlab.cern.ch/atlas/athena/-/merge_requests/55990MM offline DQ - plit PBC in 2 FEB in occupancy plots2022-08-18T21:35:15+02:00Michela BigliettiMM offline DQ - plit PBC in 2 FEB in occupancy plots- split PBC in 2 FEBs in occupancy histograms
- some cleanup
tagging DQ convenors: @sara @ytakubo- split PBC in 2 FEBs in occupancy histograms
- some cleanup
tagging DQ convenors: @sara @ytakubohttps://gitlab.cern.ch/atlas/athena/-/merge_requests/55774athanalysis fix for mc event weight variation2022-09-02T06:44:29+02:00Victor Hugo Ruelas Riveraathanalysis fix for mc event weight variationExtension fix for https://gitlab.cern.ch/atlas/athena/-/merge_requests/55737.
### Summary of original MR:
This is a "fix" for an issue I was getting running a few algorithms that used `CutFlowSvc` in `AthAnalysis`.
On the first event ...Extension fix for https://gitlab.cern.ch/atlas/athena/-/merge_requests/55737.
### Summary of original MR:
This is a "fix" for an issue I was getting running a few algorithms that used `CutFlowSvc` in `AthAnalysis`.
On the first event I get
```
CutFlowSvc ERROR Inconsistent weights and variation sizes 217 and 1
terminate called after throwing an instance of 'std::runtime_error'
what(): Inconsistent weights and variation sizes
```
I'm tagging @tadej, @khoo, @jburr, @frenner, @viruelas. I'll work on an MRE, for now the problem is demonstrated [in some analysis code](https://gitlab.cern.ch/dguest/hh4b-analysis/-/blob/e18dc5ef3de0feccf63600c76e64e2b6df1eb57f/HH4bAnalysis/share/VariableDumperConfig.py).
Roughly speaking, I set up `MainServicesConfig`, `SystematicsSvc`, `TrigDecisionTool`, `TrigEventSelectionAlg`, and call `makeEventSelectionAnalysisSequence`.
Does anyone have this working in `AthAnalysis`?Victor Hugo Ruelas RiveraVictor Hugo Ruelas Riverahttps://gitlab.cern.ch/atlas/athena/-/merge_requests/55832Sweeping !55802 from master to 22.0-mc20.
Fix obsolete scheduling of StreamBS2022-09-05T09:23:29+02:00Atlas NightlybuildSweeping !55802 from master to 22.0-mc20.
Fix obsolete scheduling of StreamBSFix obsolete scheduling of StreamBS
See merge request atlas/athena!55802Fix obsolete scheduling of StreamBS
See merge request atlas/athena!55802https://gitlab.cern.ch/atlas/athena/-/merge_requests/57139Draft: unittest test do NOT merge2022-09-30T10:10:22+02:00Marcin NowakDraft: unittest test do NOT mergeRemoved the use of the IOVSvc callback as it is not thread-safe.
Now relying solely on incidents (begin run/file)
Also removed DataHandles use to not trigger gccchecker thread safety
errors (see ATEAM-855)Removed the use of the IOVSvc callback as it is not thread-safe.
Now relying solely on incidents (begin run/file)
Also removed DataHandles use to not trigger gccchecker thread safety
errors (see ATEAM-855)https://gitlab.cern.ch/atlas/athena/-/merge_requests/56811Draft: ZDC Run 3 migration to 22.0, maintaining Run 2, but with mods to EDM2022-10-03T03:31:29+02:00Peter Alan SteinbergDraft: ZDC Run 3 migration to 22.0, maintaining Run 2, but with mods to EDMMajor change to the ZDC EDM and processing, to allow for Run 2 and Run 3 processing. Contrary to what I say in the first comment below, I based this off of 22.0, not master.
1. ZDC EDM updated such that the ZdcTriggerTowers are no lon...Major change to the ZDC EDM and processing, to allow for Run 2 and Run 3 processing. Contrary to what I say in the first comment below, I based this off of 22.0, not master.
1. ZDC EDM updated such that the ZdcTriggerTowers are no longer written out to StoreGate for Run 2. Instead, the waveform information is copied into the updated ZdcModule_v1.
a. We did *not* make a _v2 update of ZdcModule and ZdcModuleContainer, based on discussions with Attila, etc. The assumption is that we will not be using schema evolution and will rely on reprocessing older data.
2. ZdcByteStream: Run 3 unpacking (ZdcByteStreamLucrod) now added
a. Still fixed length decoding - urgent to modify to adapt to the given number of BC
b. Trigger information provided by Antonello
3. ZdcRec: Run 3 reconstruction now added
a. has a major architectural change: requiring the ZdcSums to exist before running the signal processing tool. This is because the unpacking needs a side-level ZdcModule to store the trigger information embedded in the raw data (just for the ZDC High Gain LUCROD)
b. Trigger info added to each channel, and side-level trigger sums provided by initial ZdcSum objects
4. Small modification to TrigT1CaloByteStream to remove reference to ZdcTriggerTowers (AFAIK I did not add this myself in Run 2) - might not be needed
5. ForwardRec has a first pass at ZDCRecConfig.py, to accommodate the CA framework (not tested since I don't have a testing environment for it yet)
6. ZdcConditions has the new Run 3 mapping procedures, based on a .json file in the GroupData area
7. ZdcAnalysis is still nominal - urgently needs modifications for Run 3
8. RecExPers has been modified to stream the modules and sum containers for both ESD and AOD. Could be streamlined a bit but should work and be unchanged in Run 3.https://gitlab.cern.ch/atlas/athena/-/merge_requests/57583The get methods of the ROBDataProviderSvc can be made const2022-10-17T20:13:24+02:00Christos Anastopouloschristos.anastopoulos@cern.chThe get methods of the ROBDataProviderSvc can be made constThe get methods of the ROBDataProviderSvc can be made const
This seems possible here.
We seem to have ``tbb::concurent_map`` and also a ``lock`` for the hlt implementation,
when doing kind of a lazy init from ``getRobData`` via the int...The get methods of the ROBDataProviderSvc can be made const
This seems possible here.
We seem to have ``tbb::concurent_map`` and also a ``lock`` for the hlt implementation,
when doing kind of a lazy init from ``getRobData`` via the internal ``eventCache_addRobData``
ping @tadejhttps://gitlab.cern.ch/atlas/athena/-/merge_requests/56350Draft: AthDerivation: Make sure CA-based jobs can run + disable some tests2022-10-21T11:06:08+02:00Tadej Novaktadej.novak@cern.chDraft: AthDerivation: Make sure CA-based jobs can run + disable some testsThis MR adds a few more packages to `AthDerivation` so CA-based `PHYS` can run both for MC20 and Upgrade (and probably also for MC21). This only adds O(300) compilation units so it should not increase the build too much (it has almost 20...This MR adds a few more packages to `AthDerivation` so CA-based `PHYS` can run both for MC20 and Upgrade (and probably also for MC21). This only adds O(300) compilation units so it should not increase the build too much (it has almost 20k compilation units already).
By advice from @akraszna I also introduced `ATLAS_DERIVATION_RELEASE` so we can disable tests. For now I disabled
- `EventInfo` overlay tests
- `InDetConfig` tests as they all run ~full reconstruction
- `MuonConfig` tests as they similarly run ~full reconstruction
I may look at some other tests in the next days. Before undrafting I want to confirm that this is not really tested and the CI will only confirm this does not break other projects (hopefully).
Tagging also @jcatmore and @calpigia for derivations, @npetters and @tstreble for upgrade.Tadej Novaktadej.novak@cern.chTadej Novaktadej.novak@cern.chhttps://gitlab.cern.ch/atlas/athena/-/merge_requests/54916PanTauAlgs+AthLinksSA: manual sweep from 22.0 and fix for AnalysisBase2022-11-08T09:59:30+01:00Frank WinklmeierPanTauAlgs+AthLinksSA: manual sweep from 22.0 and fix for AnalysisBaseManual sweep of !53791 from ~"22.0" to ~master.
In order for this to compile in AnalysisBase, the `getDataNonConstPtr` method had to be added to the standalone
version of `ElementLink`.Manual sweep of !53791 from ~"22.0" to ~master.
In order for this to compile in AnalysisBase, the `getDataNonConstPtr` method had to be added to the standalone
version of `ElementLink`.Frank WinklmeierFrank Winklmeierhttps://gitlab.cern.ch/atlas/athena/-/merge_requests/55892Draft: Make mis-configured ItemList an ERROR rather than WARNING2022-11-23T21:18:50+01:00Peter Van GemmerenDraft: Make mis-configured ItemList an ERROR rather than WARNINGAs discussed in Jira: ATEAM-851 this MR will cause an ERROR for mis-configured ItemLists.
However it does _not_ yet modify ~5 clients that use versioned types, as in the very past this may have been ok.As discussed in Jira: ATEAM-851 this MR will cause an ERROR for mis-configured ItemLists.
However it does _not_ yet modify ~5 clients that use versioned types, as in the very past this may have been ok.Peter Van GemmerenPeter Van Gemmerenhttps://gitlab.cern.ch/atlas/athena/-/merge_requests/58853Daily "22.0 to master" sweep: 29/11/222022-12-01T00:54:36+01:00Vakhtang TsulaiaDaily "22.0 to master" sweep: 29/11/22One merge conflict originated from !58752. As [discussed](https://gitlab.cern.ch/atlas/athena/-/merge_requests/58752#note_6236591) with @pavol on this MR, the commit was excluded from the sweep. A manual sweep will be made later on.One merge conflict originated from !58752. As [discussed](https://gitlab.cern.ch/atlas/athena/-/merge_requests/58752#note_6236591) with @pavol on this MR, the commit was excluded from the sweep. A manual sweep will be made later on.https://gitlab.cern.ch/atlas/athena/-/merge_requests/58915Draft: Trk track summary value2022-12-02T00:45:44+01:00Christos Anastopouloschristos.anastopoulos@cern.chDraft: Trk track summary valuehttps://gitlab.cern.ch/atlas/athena/-/merge_requests/58917Draft: Try to move Trk::Track summary to value2022-12-02T23:37:08+01:00Christos Anastopouloschristos.anastopoulos@cern.chDraft: Try to move Trk::Track summary to valueTry to move Trk::Track summary from ``ptr`` to ``value``.
This follows having value for the ``FitQuality`` of the TrackStateOneSurface .
It might be more tricky though
Should be lots of one-linersTry to move Trk::Track summary from ``ptr`` to ``value``.
This follows having value for the ``FitQuality`` of the TrackStateOneSurface .
It might be more tricky though
Should be lots of one-linershttps://gitlab.cern.ch/atlas/athena/-/merge_requests/59967Load more high-pt minbias is NumberOfHighPtMinBias is set too low2023-01-21T01:18:44+01:00Beojan Stanislausbeojan.stanislaus@cern.chLoad more high-pt minbias is NumberOfHighPtMinBias is set too lowShould solve issue raised in !59957Should solve issue raised in !59957https://gitlab.cern.ch/atlas/athena/-/merge_requests/55086Draft: RNTuple Storage Technology for Athena/APR2023-03-03T12:16:39+01:00Marcin NowakDraft: RNTuple Storage Technology for Athena/APRThis is an implementation of a new APR Storage Technology using ROOT RNTuple.
The eventual goal is to be able to use RNTuple in the same way as we use TTrees and TBranches now, simply by specifying the new StorageTech.
RNtuple Storag...This is an implementation of a new APR Storage Technology using ROOT RNTuple.
The eventual goal is to be able to use RNTuple in the same way as we use TTrees and TBranches now, simply by specifying the new StorageTech.
RNtuple Storage Technology can be enabled by setting this property:
AthenaPoolCnvSvc::StorageTechnology Property to ROOTRNTUPLE (now ROOTTREEINDEX)
Things not working yet:
* more complicated cases of DataVector can not be read back
* ElementLinks will not be initialized correctly when reading (no container name)
* backfilling for new dynamic attributes is disabled
* no support for std::map and std::set
Build procedure: use **asetup master--dev3LCG,latest,Athena**
packages to build:
+ Database/AthenaPOOL/AthenaPoolCnvSvc
+ Database/APR/.*
+ Database/AthenaRoot/RootAuxDynIO
+ Event/xAOD/xAODCore
Implementation details:
There is one RNTuple per original TTree, name is changed to RNT:<TreeName>Marcin NowakMarcin Nowakhttps://gitlab.cern.ch/atlas/athena/-/merge_requests/59630Draft: Create a new global instance for legacy usage of CA-based ConfigFlags2023-03-09T10:17:54+01:00Tadej Novaktadej.novak@cern.chDraft: Create a new global instance for legacy usage of CA-based ConfigFlagsCreate a new global instance for legacy usage of CA-based `ConfigFlags`. The goal is to move away from the default global instance but keep one for legacy usage (where global state is expected anyways).
The API has not changed, just a n...Create a new global instance for legacy usage of CA-based `ConfigFlags`. The goal is to move away from the default global instance but keep one for legacy usage (where global state is expected anyways).
The API has not changed, just a new `globalConfigFlags` alias is now available.
/cc @jchapman, @fwinkl, @wlampl, @christos, @mhodgkin, @jcatmore, @nstyles, @tstreble