CERN QMUL
L1Calo Software Meeting at RAL on 20 March 2001
Present: Bruce, Eric, Gilles, Murrough, Norman, Steve.Agenda
Actions from Previous Meetings
- Norman has now ordered three Concurrent CPUs similar to the one at QMW (possibly apart from 3-row/5-row connectors?)
- Norman has however been busy with the CMM and has not made much progress with the documents he has undertaken to write (VME--, event formats). The LVL1-LVL2 interface document also needs updating - once the new JetSumEt ROI is defined.
- Gilles has started writing up the CPM test plan he presented at the last meetings as a short document.
- The RAL based group has not had the time to make a start on documenting a test plan for the whole system.
Preparations for the ROIB Test
Bruce, with Normans help recently, has made good progress with the ROD ROI firmware tests. The DSS firmware was updated by Viraj to correct problems with the pulse register. With this fix, Bruce was able to send repeated trains of 400 events each containing 64 ROIs through the DSS-ROD-DSS system, reading them out and checking them. Without readout, the ROD can manage event rates up to 800kHz. So the required 100kHz ATLAS trigger rate is easily surpassed. In one overnight run no errors were detected.Some aspects of the ROD functionality remain to be implemented in firmware (eg the BUSY) or tested (eg permutations of error bits).
For the ROIB tests, we will ship our whole setup to CERN, including crate, DSS, ROD, TTCvi, TTCvx, cables, the QMW CPU (on loan) plus disk. Bruce will be there most of the two week test period; Norman for parts of two weeks, Murrough for part of one week.
Software Schedule
At the recent Birmingham meeting Murrough presented a list of software tasks for the immediate module and subsystem tests, slice tests and the final system.We need to agree a realistic schedule and decide which tasks should have the priority since we are unlikely to be able to complete them all by the start of the slice tests.
The main areas of immediate concern are the development of HDMC, eg new Parts useful for module tests and the readout system. The arrival of the PC ROS from CERN keeps slipping. For some purposes we need to read data from modules which dont have Slinks so we may need private, ROD crate DAQ style developments for that. However the balance of effort into interactive diagnostics versus the DAQ type depends on how the modules will be tested. Having better defined test strategy for the modules will help determine the best approach.
Regarding individual workplans:
- Murrough intends to continue work related to the run control and system setup, databases etc. Also perhaps development of HDMC or other hardware access library framework to be used in the DAQ.
- Bruce will continue with ROD/ROIB tests and will complete the tests of the CP slice data firmware.
- Gilles expects to work on CPM tests including developing the required HDMC support: configuration and maybe Parts?
- Steve is happy with his assigned tasks related to monitoring simulation and an overarching framework for tests.
- Bill is unfortunately unwell at the moment. We hope that he will continue work on test vectors.
Status of Test Plan Documents
Largely covered in the actions above. Much is still outstanding.Recent Work
- ROD tests: covered in the discussion of ROIB tests
- Run control/database: Murrough has started writing a document concerned with the databases we need. A new version of the Online software has just been released with an improved Information Service. This eases the use of the IS in both C++ and Java. Murrough intends to try this for prototyping IGUI panels setting run parameters etc.
- Monitoring/simulation: Steve is working on his ideas and will present them in a document.
Web and Documentation
Murrough has revamped the software web pages which are now here.The website is kept in our CVS repository. This is mainly to allow the rest of the group be able to add new pages or links to their home pages more easily.
The website also has a list of software documents which are also kept in CVS.
Calibrations
Murrough has started yet another draft document listing all the types of calibration we expect to need. So far not much detail.AOB
We had (another!) long discussion related to the use of HDMC in the DAQ context. Despite many previous intentions of Murrough and Bruce to sit down together and define the shape of a suitable hardware access library for use in the DAQ, this has not been done and is becoming urgent.Actions
- Murrough and Bruce: agree development plan for HDMC/hardware access library for the DAQ
- Steve: document ideas for monitoring/test framework
- Norman: finish various documents...
- Murrough: finish run control related documents for review
- Gilles: finish CPM test plan document
Next meeting
No date for the next meeting has been set. It will be after Easter and probably after Bruce returns from well deserved holidays...Last updated on 20-Mar-2001 by Murrough Landon