Software phone conference on 14 June 2004

CVS server

Our CVS server is now at CERN and seems to be working OK now. The nightly builds have restarted.

Only the l1calo area was moved. The firmware repository and obsolete software are still at RAL. These are accessible (only) via the RAL AFS if required.

PPM software

Florian has made a lot of updates to ppmServices and ppmTests. Apart from the database work which he has been using locally, he has also started on readout of the rate meter and histograms. He intends to fill ROOT histograms. These should be published in the IS server. The correct version of ROOT (3.10.2) needs to be installed at Heidelberg in our common location: /opt/share/utilities/rootVersions

Kambiz reported on a discussion with Bruce and Murrough about the best way to access the ASIC registers (and some others). These need a lengthy arbitration - too long to keep hoplding the VME bus for a single cycle. Instead, for registers there will be a special protocol to be encapsulated in a new ModuleRegister32 subclass. This will use the top two bits of each register as flags to request read/write transfers on one cycle and check for their complete on a second cycle. Memories will be accessed via block transfer with the protocol explicit in higher level software.

For access to the TTCrx Florian and Kambiz will try to use the common approach adopted for the other modules.

Bruce mentioned that the default VMEBusCERN implementation was now the programmed IO version. The old A24 memory mapped implementation is renamed. Florians A32 memory mapped bus will be VMEBusCERNa32.

Other modules

In Stefans absence, Murrough reported that the new jemServices and jemTests software for JEM1 had been committed last week. The old JEM0 software was preserved as separate jem0Services and jem0Tests packages. One or two other changes would also be needed if the JEM0 was really going to be used again.

Bruce mentioned that the 9U ROD and its firmware were expected by the end of July.

Monitoring

Adrians histogram display and monitoring framework packages were added to CVS last week. He is about to add documentation for them. The packages contain an example monitoring program.

To use the new monFramework package requires a patch to one of the header files of the online oh package. An official patch is expected from the online group, but not immediately. In the mean time, Adrian will distribute his patch.

Adrian has also been looking at the monitoring requirements document and suggested that in addition to the event stream we should create a second data stream with statistical information. However although better treatment of the existing "hardware status information" was desirable, it was felt that work on decoding and monitoring the event data was a higher priority.

It was suggested that we should have a brainstorming on treatment of the statistical type information and also on byte stream decoding at the Stockholm meeting.

One critical requirement for developing the byte stream decoding for the test beam is that we have a decision on exactly which ROD formats will be used. It was proposed that the next hardware phone conference (Thursday 17th) was a suitable forum to take this decision.

ROD Crate DAQ

Murrough has made some further progress in understanding how to use our software in the ROD crate DAQ controller framework. There are still some outstanding questions for which input is needed from the ROD crate DAQ support team. It is still proposed to attempt some tests during the week of 21 June at RAL with the ROD crate DAQ software.

AOB

• Bruce raised the issue of our event fragment interface. We invented this for greater generality. At the moment Adrians new monFramework package doesnt use it. We agreed that it should. We may also want a further layer between this and byte stream decoding. This can be discussed at Stockholm.
• Adrian suggested we should define a naming convention for storing our histograms in the IS server.
• Kambiz said that in a few weeks Pavel will start working with Florian on PPM online software, perhaps generating test vectors.

Next meeting(s)

There will be a software discussion/brainstorming meeting at Stockholm on Wednesday 30 July, starting at 9:30(?).