Re: subArray processing

[Carl Lionberger <calionberger@lbl.gov>]

Hi Hinko,

I agree that MALM should be 16 in this case.  My question is what client 
you are using to show the values you are dissatisfied with.  It seems to 
me that it is mishandling the metadata for the subarray record, or 
perhaps the metadata from the subarray is not appropriate.  Perhaps it 
is reading NORD, which would be appropriate for a waveform, but not a 
subarray record.
Carl

Hinko Kocevar wrote:

> Hello,
>
> I'm trying to use subArray record type to display each array element at
> index 0 .. n (n == 10; my case) in a separate PV.
>
> record(bi, "$(DEVICE)ACQUIRE") {
> 	field(SCAN, "I/O Intr")
> 	field(DTYP, "asynUInt32Digital")
> 	field(FLNK, "$(DEVICE)MONITOR PP")
> 	field(INP, "@asynMask($(PORT), 200, 0x1, 1.0)")
> }
>
> record(waveform, "$(DEVICE)MONITOR") {
> 	field(DESC, "A B C D X Y Q SUM CX CY")
> 	field(DTYP, "asynInt32ArrayIn")
> 	field(FLNK, "$(DEVICE)A PP")
> 	field(INP, "@asyn($(PORT), 200)")
> 	field(NELM, "16")
> 	field(FTVL, "LONG")
> }
>
> record(subArray, "$(DEVICE)A") {
> 	field(INP, "$(DEVICE)MONITOR")
> 	field(INDX, "0")
> 	#field(MALM, "16")
> 	field(MALM, "1")
> 	field(NELM, "1")
> 	field(FTVL, "LONG")
> 	field(FLNK, "$(DEVICE)B PP")
> }
>
> record(subArray, "$(DEVICE)B") {
> 	field(INP, "$(DEVICE)MONITOR")
> 	field(INDX, "1")
> 	#field(MALM, "16")
> 	field(MALM, "1")
> 	field(NELM, "1")
> 	field(FTVL, "LONG")
> 	field(FLNK, "$(DEVICE)C PP")
> }
>
> Above are the records that I use. But the record A and B do not work as
> expected - A and B both display MONITOR element at index 0. See below:
>
> LIBERA01:SA:MONITOR            2009-06-03 10:31:25.800912 16 4464720
> 4614628 5834728 7518137 682839 -1902545 -816500 22432214 9866483 657703
> 1973108 3288513 4603918 5919323 7234728 8550133
> LIBERA01:SA:A                  2009-06-03 10:31:25.800912 4464720
> LIBERA01:SA:B                  2009-06-03 10:31:25.800912 4464720
> LIBERA01:SA:MONITOR            2009-06-03 10:31:25.900912 16 4466474
> 4610680 5834452 7525126 688647 -1905988 -817173 22436734 9866498 657704
> 1973111 3288518 4603925 5919332 7234739 8550146
> LIBERA01:SA:A                  2009-06-03 10:31:25.900912 4466474
> LIBERA01:SA:B                  2009-06-03 10:31:25.900912 4466474
> LIBERA01:SA:MONITOR            2009-06-03 10:31:26.000912 16 4470273
> 4605824 5816126 7515386 697287 -1896477 -818335 22407610 9866513 657705
> 1973114 3288523 4603932 5919341 7234750 8550159
> LIBERA01:SA:A                  2009-06-03 10:31:26.000912 4470273
> LIBERA01:SA:B                  2009-06-03 10:31:26.000912 4470273
>
> Now, if remove all the lines with:
> 	field(MALM, "1")
> 	field(NELM, "1")
> in records A and B, and add this instead:
> 	field(MALM, "16")
>
> records A and B work as expected, but they are 16 fields long - I would
> like to see them 1 element long as the really are (one 32 bit integer
> value).
>
> LIBERA01:SA:MONITOR            2009-06-03 10:30:17.280912 16 4472801
> 4610293 5819417 7532655 701680 -1900157 -824336 22435169 9856373 657029
> 1971086 3285143 4599200 5913257 7227314 8541371
> LIBERA01:SA:A                  2009-06-03 10:30:17.280912 16 4472801 0 0
> 0 0 0 0 0 0 0 0 0 0 0 0 0
> LIBERA01:SA:B                  2009-06-03 10:30:17.280912 16 4610293 0 0
> 0 0 0 0 0 0 0 0 0 0 0 0 0
> LIBERA01:SA:MONITOR            2009-06-03 10:30:17.380912 16 4465950
> 4612662 5830285 7532898 692675 -1906287 -823451 22441797 9856388 657030
> 1971089 3285148 4599207 5913266 7227325 8541384
> LIBERA01:SA:A                  2009-06-03 10:30:17.380912 16 4465950 0 0
> 0 0 0 0 0 0 0 0 0 0 0 0 0
> LIBERA01:SA:B                  2009-06-03 10:30:17.380912 16 4612662 0 0
> 0 0 0 0 0 0 0 0 0 0 0 0 0
> LIBERA01:SA:MONITOR            2009-06-03 10:30:17.480912 16 4467876
> 4612356 5839784 7524717 685678 -1906164 -814509 22444735 9856403 657031
> 1971092 3285153 4599214 5913275 7227336 8541397
> LIBERA01:SA:A                  2009-06-03 10:30:17.480912 16 4467876 0 0
> 0 0 0 0 0 0 0 0 0 0 0 0 0
> LIBERA01:SA:B                  2009-06-03 10:30:17.480912 16 4612356 0 0
> 0 0 0 0 0 0 0 0 0 0 0 0 0
>
> How can one implement single element PV in subArray record?
>
> Thank you,
> Hinko
>
>  

--
Carl Lionberger
Control System Software Engineer
Berkeley National Laboratory (LBNL)
(510) 486-7503