Minutes from the 06/07/05 and 06/14/05 LCLS SLC IOC meetings: (1) Updated Task List: Task Who When ------------------------------------------ --------------- ---------- * VDCT Standards Kristi June 21 * BSAC informal design review Debbie, Diane June 22 * Modelling issues/task meeting Many June 28 * Magnet EPICS func reqts review Kristi June 30 * Port CMLOG client to RTEMS RonM, Steph June 30 * BSAC design review Debbie, Diane Mid July * ICDs - EVR (w/Dayle),BPM (w/Till) Debbie, Diane Late July * ICD - Toroid and GADCs (w/Doug) Debbie, Diane Late July * Magnet SLC design review Kristi July * Magnet EPICS design review Kristi July * PNET timing diags review Steph July * BPM SCP and SLC DB for BPM testing MikeZ July * Mgnt SCP and SLC DB for Mgnt testing ????? July * Decide on SCP wire-scanner Controls Group July * BSAC code/test finished Debbie, Diane Aug 17 * Magnet code/test finished Kristi Aug 17 * PNET timing diags code/test finished Steph Aug 17 * SCP and SLC DB for Oct test stand ????? Oct * Support test stand setup Steph Oct Loose End Tasks Who When ------------------------------------------ -------------- ----------- * Fix database conversions on Linux Diane June 21 * Timing of 10 sec update not right RonM June 21 * Get newer version of VDCT on Linux Steph June 21 * CSTR/ASYNC code review Steph June 30 * Change exec to start correct threads Diane Early July * SLC restart bug Diane Aug 17 * Wrong IPL time when not booted from SCP RonM Wait on reqt * Thread priorities Steph Aug 17 * slcCmlog change to honor stop flag Steph Aug 17 * Update specs All As needed * Update how-to-add-a-new-micro Steph Aug 17 * Update record naming conventions All As needed TBD: * Increase max number of micros * Second production proxy? * Error metering * VMS changes to support SLC IOC * LCLS SLC modelling database setup (Mark Woodley?) * "Domain" split (2) Basic Services: Loose ends task list added (see above). Ron corrected PARANOIA so crate verification not sent to IOC. Ron is having difficulties porting the CMLOG client to RTEMS. It uses a package called ACE. CMLOG uses an old version of ACE (actually uses a subset of the entire ACE package) but only the new version is available for RTEMS. Ron has 3 options: * get the new version and upgrade CMLOG to use it * get the OS.h (contains platform-specific macros) from the new version of ACE and use it to help port the old version of ACE to RTEMS * forget about putting the CMLOG client on RTEMS and instead change the iocCmlogHdlr to send the ASCII cmlog tags to a process on unix that forwards the tags to the cmlog server. (3) Beam Synchronous Acq and Control (BSAC): Diane and Debbie finished the BSAC func reqts and had a review: http://www.slac.stanford.edu/grp/lcls/controls/global/sw/slc_ioc/requirements/BSACFacilityReqs.htm Informal design review with Bob on June 22. A design review with the rest of the group will be later and will NOT involve a formal presentation. MikeZ estimates 1 week to upgrade SCP to use EGU values and some way to identify the stepper motor instead of camac count values and control words for the SLC IOC. Decision to use SCP wire scanner package will be made in July. The design of the BSAC goes forward assuming no beam-synch'ed control. It's judged that this feature can be added later with minimal changes. BSAC will process the WIRE_PREP message but will only do the buffered acquisition part. Till must provide TMIT, along with X and Y, as ai records used by BSAC. Dayle must provide parts of the proper timing pattern from the EVR pattern pipeline as records used by BSAC. Processing of BSAC records will be started by Till, either via scanIORequest in his device support or by FLNK from Till's last processed record for the beam pulse. Note that Till will plan on his own diagnostic records which will be independent of BSAC and will not be a part of any SCP message. ICDs are needed to document these assumptions. Most of the BSAC processing (except timing) can be done on Linux using simulated X, Y, TMIT, and pattern records. BSAC record names are being decided - the LCLS record naming conventions will be updated. (4) Visual Database Configuration Tool (VDCT): Kristi wants to develop VDCT standards: * Standard macro name recommendations * What sort of text to add to a VDCT diagram - title and macro definitions - description of external links - brief subroutine description for subroutine records - extra description for complicated sets of links * Rule of thumb for size of diagram (normally should fit on one screen) Kristi corrected the LCLS setup files for VDCT. We need the newer version of VDCT available to the SLAC linux servers. (5) Magnet Service: Kristi finished the func reqts for the SLC side of the magnet service and had a review: http://www.slac.stanford.edu/grp/lcls/controls/global/sw/slc_ioc/requirements/mgnt_funcReqts_V2.html She is working on the func reqts for the EPICS side now and should have another review (probably after Patrick gets back). LCLS will be adding new magnets in the middle of the existing sectors (LI21 to LI30 and maybe in the BSY too). The new magnets will be mixed in with the old magnets. The new magnets will be assigned to their own "micros" with names like LA21 to LA30. So to get the correct order of magnets for something like modelling, you can't get all of LI21's magnets and then append LA21's magnets, but instead have to intersperse them depending on Z position. This requires potentially significant changes to the Alpha. There will be more discussion with ESD folks on this issue plus other SLC-modelling-related tasks in the next few weeks. Kristi continues to work on the PS low-level records (name changes, adding more records). She continues to rtemize the device/driver support. In the future, she plans on adding more records to support non-SLC high-level apps. She is not sure what is included in the XML files and what should be records on the IOC. She would like to know more about plans for names and types of information in RDB. Magnet record names are being decided - the LCLS record naming conventions will be updated.