Re: OSI Compliant VME Support

[Matt Rippa <mrippa@gemini.edu>]

Sorry I wasn't clear. Andrew's XYCOM code includes features we have at Gemini. Of course we'll be using Base R3.14 to make them OSI-compliant.

-Matt

On Wed, Feb 24, 2016 at 8:26 AM, Matt Rippa <mrippa@gemini.edu> wrote:

Hi Michael and Andrew,

I will need compatibility with the Base-3.13.x. Seems the best outcome is a single common codebase. Michael, would that work for you?

-Matt

On Wed, Feb 24, 2016 at 7:47 AM, Andrew Johnson <anj@aps.anl.gov> wrote:

Hi Michael,

On 02/23/2016 10:38 PM, Michael Davidsaver wrote:
> https://github.com/mdavidsaver/xycomioc
>
> Support for 220, 240, and 566 cards.  Tested with RTEMS and vxWorks.
> Can't remember how feature complete.  Other than checking that it builds
> just now, I haven't looked at it since 2010.  This code holds some
> sentimental value to me as it's one of my first epics drivers, and the
> only one I found a good name for.

Nothing against your drivers, but Matt should understand that they were
not derived from the xycom drivers that were included with Base-3.13.x
and earlier. IIRC the original xycom digital drivers had a slightly
unexpected mapping between the signal address and I/O channels (S0 is
not port 0 bit 0). The xy240 driver also has a background thread that
emulates I/O Interrupt scanning.

I suspect Matt has existing applications that expect the original
capabilities, so compatibility may be important to him.

- Andrew

--
There are only two hard problems in distributed systems:
  2. Exactly-once delivery
  1. Guaranteed order of messages
  2. Exactly-once delivery
 -- Mathias Verraes