CAMAC Front End Fix List

Date
entered
Date
completed
Assigned
to
Description
04-25-01 04-26-01 RN Can't talk to 290s using H32.

Update

H32 was using a retry of 1. This was too low. The SLD driver was changed to clip retries to 15 or higher.

H32 was modified to default to the NCAMAC front-end.

04-25-01 04-25-01 RN "tScanLam" is confused in reporting LAMs.

Resolution

When I updated tScanLam to do half the work every second, I missed a spot were the crate was being miscalculated.

04-20-01   RN/PK Fix block transfer on MI.

Right now, two things remain broken. The SLD needs to be able to handle the signals on the BTR line. Cleaning up the signals buys us some time, but eventually (as things age) we may start seeing problems again. Paul is working on this.

The other problem is something needs to be done to recover the SLD when it stops reporting interrupts (other than rebooting the front-end.)

Update

Paul has updated the firmware in the SLD. It is installed in the test front-end. The driver needs to be slightly modified (since the new code handles more details.)

04-20-01   RN The H5 application needs to support the SLD.
04-20-01 07-31-01 RN The D21 application no longer works. It doesn't reliably send the CNAFs on the link. It is not clear whether CLIB has changed, D21 has changed, or the UPLINK task has changed.

Update

The problem was due to several things. The first is that if multiple crates were being sent commands and one of those crates returned a timeout, the rest of the entries would be garbage.

The second problem was that several tasks that use UPLINK were specifying retries of 1. The SLD driver was updated to set the retry limit to 15, if the caller specified a retry less than 15.

Another bug was fixed. D21 sends up to 20 CNAFs. There may be holes in the list of CNAFs. Unfortunately, UPLINK was merging the results into a smaller list, so if any gaps were sent to the front-end, the results were screwed up.

The extended status information is also returned now.

04-20-01   RN Fix BOOSTR BLMs. MOOC implements the FTD timing in software, which is too coarse for BOOSTR BLMs. Need to modify the MADC support to set up a list when non-standard FTDs are used.
04-20-01   GJ/RN Fix GAS alarm support. GAS alarms are not going to be handled as in the old front-end. Instead we'll let MOOC scan the alarms like it does with other devices.

Update

Support for alarms has been added, but not fully tested. This support was done through the MOOC class (i.e. we're not downloading the alarm blocks into the GAS modules.)

Also, support was added to tScanLam to search for GAS modules when interfacing with AERBOOT.

04-20-01     C181 devices aren't fully operational. The driver needs to be fixed. But since no further information is available as to the missing functionality, no more progress can be made.
04-20-01 06-18-01 RN There was a property of the old *:BTRON devices that would toggle the offline/online status of every crate. We need to implement this.
04-20-01   RN Implement the functionality of the *:PIOX devices.

Update

*:PIOX devices are now "supported". They don't generate alarms since the return value is fudged, but they are now off Lin's error list. We need to add an "equipped crates" list to these front-ends.

04-20-01     The TIMSPY device doesn't appear to be operating the same as the MB-II version.
04-20-01 04-20-01 RN Add LAM reporting support to 479 modules.
04-20-01   GJ Add LAM reporting support to 080 modules.
04-20-01     Add LAM reporting support to 200 modules.
04-20-01     Add LAM reporting support to 193 modules.
04-20-01     Add LAM reporting support to 181 modules.
04-20-01 04-26-01 RN Add LAM reporting support to 276 modules.
04-20-01     Add LAM reporting support to 145 modules.
08-14-01   RN C23 is having problems collecting data from 290s. It seems that the 290s occasionally miss a TCLK event and return stale data (even though a DIO_STALE isn't returned to the application.)
08-14-01   RN Elvin Harms reported problems reading I:IBEAMS and I:TOR521 from SDA. These devices had no problem being plotted. They couldn't be accessed through the DAEs.

Richard Neswold

Security, Privacy, Legal