Experimental Physics and
Industrial Control System

[email protected] (Ned Arnold) · Mon, 11 Jul 94 13:07:25 CDT

> One possible solution would be to tweak the driver a bit and create 5
> device-support packages (32 bits in & no bits out, 8 bits in & 24 bits
> out, ... no bits in and & 32 bits out), but that seems kind of ugly.
> 
> The other options seem even uglier, and much more prone to error in the
> configuration process.
> 
> I'm probably missing the obvious here (now that's something I seem to excel
> at!).  Any ideas would be appreciated.
> 
> -bill
> 

I'm always an advocate for making the interface to the application developer
as easy as possible, so I agree that the 5 DTYP's are not very pretty. 

How about having the driver allocate whether a byte is input or output
depending on the first bit request for that octet. For example, if Bit 3 is
specified in an input record (bi, mbbi), have the driver keep track that 
bits 0 thru 7 are inputs. If an output record later requests bit 7, issue an
error message that this cannot be accommodated. 

This way, the application person doesn't have to worry about anything unless
he/she has assigned bits in contradiction to the hardware capabilities, and
then an error will be printed.

To keep track of this, you will have to allocate some scratch space for each
byte of each possible board. This does not seem like alot of memory.

	Ned Arnold
        APS

Experimental Physics and Industrial Control System

Experimental Physics and
Industrial Control System