OK now I remember why we moved the WFD_PATREG clears into equip 32.
It was to stop exactly the behavior you described. After a equip 42
error the acq bails and the clears don't happen. As you noticed this
causes problems since you just keep adding bits to the pat reg.
I don't like adding another equipment. I think we should go back to
the way it was before(everything in equip 32) and get the equipment
32s out of the ID=2 buffers. After all this sort of thing will be a
general problem in the future. If other large stuff ever gets added
into the data stream we have more then just equip 32 and 42 in the
ID=2 buffers. Chris O has confirmed that the equip 32s are in the
ID=2 buffer. The more we spread the same functionality around in
different equipments the more likely people are to make errors.
If people really object to having to get equip 32 out of the ID=2
buffers then I suggest you make a new equipment for the SPU read and
leave all of the clears in equip 32. After all everything else in
there is a clear anyway. As to the argument that it will break
software... Well we already made it "disappear" for a month and no
one even noticed.
As to the the hardware problems you are having. They are most likely
completely unrelated to the equip 32 problem. They are almost
certainly related to the hardware changed you made on microvax #2. I
guess you moved the SM4 stop master(am I right?). The behavior you
describe( equip 42 errors and 999 errors during WFD initialization) are
caused by trying to readout out and trying to set the thresholds when
when the WFDs aren't stopped respectively. Probably when you moved
things (or put in the Tees) you were just living on the edge somehow.
The first thing to check is if SM4 is *ever* getting stopped. If the
lights are always green your job is pretty straightforward. If they
are only missing sometimes the job is harder. Be careful to remember
the fact that the 2 stop masters are connected to each-other, it can
cause a lot of headaches trying to debug these problems.
-Chris