ITS: staggering#15188
Conversation
Signed-off-by: Felix Schlepper <[email protected]>
Signed-off-by: Felix Schlepper <[email protected]>
Signed-off-by: Felix Schlepper <[email protected]>
Signed-off-by: Felix Schlepper <[email protected]>
Signed-off-by: Felix Schlepper <[email protected]>
Signed-off-by: Felix Schlepper <[email protected]>
Signed-off-by: Felix Schlepper <[email protected]>
Signed-off-by: Felix Schlepper <[email protected]>
Signed-off-by: Felix Schlepper <[email protected]>
Signed-off-by: Felix Schlepper <[email protected]>
Signed-off-by: Felix Schlepper <[email protected]>
Signed-off-by: Felix Schlepper <[email protected]>
Adapt ITS/MFT CTF machinery to staggered data
Fix compilation of ALICE3 tracking with staggering
Signed-off-by: Felix Schlepper <[email protected]>
Signed-off-by: Felix Schlepper <[email protected]>
Signed-off-by: Felix Schlepper <[email protected]>
Signed-off-by: Felix Schlepper <[email protected]>
Signed-off-by: Felix Schlepper <[email protected]>
Signed-off-by: Felix Schlepper <[email protected]>
Signed-off-by: Felix Schlepper <[email protected]>
Signed-off-by: Felix Schlepper <[email protected]>
Signed-off-by: Felix Schlepper <[email protected]>
|
REQUEST FOR PRODUCTION RELEASES: This will add The following labels are available |
Signed-off-by: Felix Schlepper <[email protected]>
|
Error while checking build/O2/fullCI_slc9 for 507223a at 2026-04-01 07:39: Full log here. |
New fastmultEst and related reworking for UPC
Signed-off-by: Felix Schlepper <[email protected]>
|
Error while checking build/O2/fullCI_slc9 for 92cd379 at 2026-04-06 00:20: Full log here. |
|
@f3sch fine for me. |
Signed-off-by: Felix Schlepper <[email protected]>
Signed-off-by: Felix Schlepper <[email protected]>
Signed-off-by: Felix Schlepper <[email protected]>
Signed-off-by: Felix Schlepper <[email protected]>
|
Error while checking build/O2/fullCI_slc9 for 9995505 at 2026-04-13 00:18: Full log here. |
|
Merging after fixing naming conflict. The CI's were passed except the conflict with the QC, whose relevant PR will be merged after this one (was tested locally by @mpuccio). |
* ITS: staggered tracking Signed-off-by: Felix Schlepper <[email protected]> * ITS: various fixes also for GPU Signed-off-by: Felix Schlepper <[email protected]> * ITS: fix vertexer and move new types Signed-off-by: Felix Schlepper <[email protected]> * ITS: format Signed-off-by: Felix Schlepper <[email protected]> * ITS: account for layer ROF bias in tracker Signed-off-by: Felix Schlepper <[email protected]> * ITS: sort tracks in time by lower edge Signed-off-by: Felix Schlepper <[email protected]> * ITS: ensure mc labels are nullptr Signed-off-by: Felix Schlepper <[email protected]> * ITSMFT: account for possible delay of received ROFs Signed-off-by: Felix Schlepper <[email protected]> * ITS: staggered STF decoder Signed-off-by: Felix Schlepper <[email protected]> * ITS: fix track time-assignment Signed-off-by: Felix Schlepper <[email protected]> * ITS: output vertices Signed-off-by: Felix Schlepper <[email protected]> * ITS: add macro to check staggering in data Signed-off-by: Felix Schlepper <[email protected]> * Adapt ITS/MFT CTF machinery to staggered data * Fix compilation of ALICE3 tracking with staggering * ITS: modify staggering macro Signed-off-by: Felix Schlepper <[email protected]> * ITSMFT: runtime staggering option Signed-off-by: Felix Schlepper <[email protected]> * ITSMFT: fix instantiation in namespace Signed-off-by: Felix Schlepper <[email protected]> * ITS3: fix compilation Signed-off-by: Felix Schlepper <[email protected]> * Raw,CTF: add option to specify base cache dir for remote files Signed-off-by: Felix Schlepper <[email protected]> * ITS: tracking same as dev Signed-off-by: Felix Schlepper <[email protected]> * ITS: add back datastreams Signed-off-by: Felix Schlepper <[email protected]> * ITSMFT: improve logging Signed-off-by: Felix Schlepper <[email protected]> * ITS: add rofs for vertices back Signed-off-by: Felix Schlepper <[email protected]> * add copyright to macro Signed-off-by: Felix Schlepper <[email protected]> * ITS: hide print functions for device code Signed-off-by: Felix Schlepper <[email protected]> * ITSMFT: add shim file for alpide param Signed-off-by: Felix Schlepper <[email protected]> * try to fix macro compilation Signed-off-by: Felix Schlepper <[email protected]> * Avoid wildcarded subspecs in Digit/ClusterWriter * ITS: fix rof lut to work properly with added errors Signed-off-by: Felix Schlepper <[email protected]> * Fix/add some staggering options * Add ITS/MFT staggering options to dpl-workflow.sh To activate ITS or MFT staggering in the topology generation, export ITSSTAGGERED=1 or MFTSTAGGERED=1 respectively * ITS: try fix for QC Signed-off-by: Felix Schlepper <[email protected]> * ITS: fix ROFLookpTables warning Signed-off-by: Felix Schlepper <[email protected]> * ITS: fix tracklet formatting Signed-off-by: Felix Schlepper <[email protected]> * ITS: set BCData properly for ROFs Signed-off-by: Felix Schlepper <[email protected]> * ITS: remove deprecated settings Signed-off-by: Felix Schlepper <[email protected]> * ITS: fix cluster label access for non-staggered Signed-off-by: Felix Schlepper <[email protected]> * ITSMFT: fix staggering wfx option for digit-writer-workflow Signed-off-by: Felix Schlepper <[email protected]> * Fix loop condition for ITS tracking layers * Fix/add some staggering options * Add ITS/MFT staggering options to dpl-workflow.sh To activate ITS or MFT staggering in the topology generation, export ITSSTAGGERED=1 or MFTSTAGGERED=1 respectively * ITSMFT: fix staggering wfx option for digit-writer-workflow Signed-off-by: Felix Schlepper <[email protected]> * Fix loop condition for ITS tracking layers * Make ITS vertex messageable * remove unused variable * Add/fix staggering options to all workflows reading ITS,MFT clusters To pass the sim-challenge test. W/o this option even <workflow> -h leads to a crash. Strictly speaking, one could use in the DPLAlpideParamInitializer::isITSStaggeringEnabled and DPLAlpideParamInitializer::isMFTStaggeringEnabled a test ic.options().hasOption(stagITSOpt) and ic.options().hasOption(stagMFTOpt) before testing the option itself. But better to have an explicit detection of missing staggering option. * ITSMFT: fix digit reader Signed-off-by: Felix Schlepper <[email protected]> * Remove leftover NROFs configurable from dpl-workflow.sh * ITS: fix time assignments Signed-off-by: Felix Schlepper <[email protected]> * ITS: fix degenerate LSE for matrix solving Comparing the output of dev and this PR, I saw plently of cases where the system of equation was fully degenerate and produced to different floating instructions and compiler optimizations slightly different results. The solution is to discard the vertex cand. if the LSE becomes degenerate as not to produce non-sense solutions. Signed-off-by: Felix Schlepper <[email protected]> * ITS: fix macro Signed-off-by: Felix Schlepper <[email protected]> * MFT: fix track writer Signed-off-by: Felix Schlepper <[email protected]> * ITS: fix gpu compile due change in vertexer types Signed-off-by: Felix Schlepper <[email protected]> * ITS: move lookup table creation to proper place Signed-off-by: Felix Schlepper <[email protected]> * Move FastMultEstimation to ITS tracking library * ITS: add containedIn to TS Signed-off-by: Felix Schlepper <[email protected]> * ITS: fix vertexer Signed-off-by: Felix Schlepper <[email protected]> * ITS: improve STFDecoder&Clusterer error messages and account for delay longer that ROF Signed-off-by: Felix Schlepper <[email protected]> * Implement new kind of multiplicity mask * Adapt GPU code to the new mult mask * ITS: finalize tracking code Signed-off-by: Felix Schlepper <[email protected]> * ITS: remove deltaRof for vertexer Signed-off-by: Felix Schlepper <[email protected]> * ITS: report current timeslice Signed-off-by: Felix Schlepper <[email protected]> * Vertex: also print time error Signed-off-by: Felix Schlepper <[email protected]> * ITS: speedup vertexer Signed-off-by: Felix Schlepper <[email protected]> --------- Signed-off-by: Felix Schlepper <[email protected]> Co-authored-by: shahoian <[email protected]> Co-authored-by: Maximiliano Puccio <[email protected]>
| std::array<int, 2> mTracklets = constants::helpers::initArray<int, 2, constants::UnusedIndex>(); | ||
| std::array<int, nLayers> mClusters = constants::helpers::initArray<int, nLayers, constants::UnusedIndex>(); | ||
| std::array<int, NLayers> mClusters = constants::helpers::initArray<int, NLayers, constants::UnusedIndex>(); | ||
| TimeEstBC mTime; |
There was a problem hiding this comment.
mTime is not initialised in the default ctor.
| vert1.getTimeStamp().setTimeStampError(594); | ||
| vertices.push_back(vert1); | ||
| table.update(vertices.data(), vertices.size()); | ||
| const auto view = table.getView(); |
| static constexpr float DEFROFLengthTrig() | ||
| { | ||
| // length of RO frame in ns for triggered mode | ||
| return N == o2::detectors::DetID::ITS ? 6000. : 6000.; |
There was a problem hiding this comment.
Why are both values same?
There was a problem hiding this comment.
Because, although these are different detectors, they have the defaults.
|
|
||
| std::vector<PairCandidate> k0Cands; | ||
| for (int iPos{0}; iPos < (int)posPool.size(); ++iPos) { | ||
| const auto pos = posPool[iPos]; |
There was a problem hiding this comment.
pos shadows an outer variable.
This is to run the CI and possibly to run tests at P2.