Discussion on storage of measurement data (i.e. not yet analyzed): We have a 2 GB partition, which is both accessible from CESR29 and from lnx180c. We need to set up the file structure so that we can store the measurement data in a way that would be similar to how this is done for the CBPM system (Mark and Eugene will look into this). This partition will serve as a mailbox, from where data will be moved to the larger storage. Devin will see how to automate this process. Hubert can help with the header routines for file name definition.
Discussion on raw data transfer to lnx180c: Plan is to test the BSM read-out software on lnx180c, whereby data is taken from one of the BSMs in CESR. Withtin one and the same application (in this case the BSM one), control is to be done using the Xbus and the large data transfer with ethernet. For the BSM we will possibly move toward control done by the ethernet as well, but for the BPM most likely not due to network latency.
In terms of libraries, Jamie has been working on adding communications software to the Beam Instrumentation Support library. Beneath this layer one can find the MPMnet and CBInet. We will also need another layer in which one would have a library containing the socket communication stuff. The naming convention proposed here is:
cbi_ , cbpm_ , mpm_ etc..
Under cbi_ one may define cbi_alloc_xxx , cbi_net_xxx, cbi_constant_xxx , cbi_dsp_xxx etc
RTEMS is avalaible on lnx180c; Tim mentions that he has placed info on the TWIKI page about this.
Mark suggest that Devin should look into another processor, as we will most likely soon run out of capicity with lnx180c (e.g. it is planned to do large amounts of bunch-by-bunch fitting, which would slow down lnx180c considerably).