Re: mbbi record ignores mask?

["Johnson, Andrew N. via Tech-talk" <tech-talk at aps.anl.gov>]

As Mark says, the MASK field is set by the record but really belongs to the device support. Some supports may want to shift it (by some signal offset related to the I/O address) but others won’t, so it’s up to the device support to modify and use it if it applies. Could be better documented, I’d be very happy to accept a patch to the mbbiRecord.dbd.pod file from the 3.15 branch if someone wants to contribute anything.

- Andrew

On Sep 15, 2020, at 10:47 AM, Mark Rivers <rivers at cars.uchicago.edu> wrote:

Ø  The mbbi record in the couple of versions of base that Ive checked (3.15.5 and 3.15.7) don’t seem to use the MASK field.

 

I believe that is the responsibility of the device support.  asyn device support handles the mask.

 

Mark

 

 

From: Tech-talk <tech-talk-bounces at aps.anl.gov> On Behalf Of Eric Norum via Tech-talk
Sent: Tuesday, September 15, 2020 10:45 AM
To: Johnson, Andrew N. <anj at anl.gov>; tech-talk at aps.anl.gov
Subject: mbbi record ignores mask?

 

This is one of those things that seems impossible to have been wrong for so long that it makes me wonder if I’m misinterpreting what I’m reading.

The mbbi record in the couple of versions of base that Ive checked (3.15.5 and 3.15.7) don’t seem to use the MASK field.

 

wenorum@xildev5 223> pwd

/eda/epics/R3.15.7/base/src/std/rec

wenorum@xildev5 224> grep -n mask mbbiRecord.c

140:    if (prec->mask == 0 && prec->nobt <= 32)

141:        prec->mask = ((epicsUInt64) 1u << prec->nobt) - 1;

wenorum@xildev5 225> 

— 

Eric Norum

eric at norum.ca

-- 

Complexity comes for free, simplicity you have to work for.