On Wed, 27 Apr 2011 07:45:53 -0400 Peter Relson <[email protected]> wrote:
:>>Is it your statement that the only "supported" use of this routine is to :>fail :>>the SVC call, not to do alternate processing? :>Yes, more or less, that is my statement. The routine may do anything it :>wants :>that is related to failing the SVC call. Good thing that I didn't realize that or thought outside the box. :>>has anyone ever submitted requirements to formalize the way :>>it really gets used? Thereby securing against future changes that might :>>break the unsupported feature? :>Not to my knowledge. And, yes, submitting such a requirement would be a :>good :>thing to do. There's not much we even could do to break the "feature". And :>I'm :>sure we would not intend to do so. But having it recognized and documented :>is always good. :>In my mind, any SVC front-ending discussion (whether screening or via :>SVCUPDTE) :>should center on "why do you need to front-end the SVC". I suspect that in :>a :>lot of cases it is because the SVC processing does not provide a suitable :>exit. :>Would not it be better for the system to provide exits rather than have :>you :>front-end SVC's? The same is true for PC's. I suspect that much of the :>answer :>lies with practicality and time-of-availability -- the SVC mechanisms are :>available now and who knows how long it would take to get what you need in :>terms of exits. Or the SVC does not do what is needed in the specific environment. Imagine if CICS instead of creating the DFHSC structure, SVC screened GETMAIN and internally did the work? Instead of DFHFC, SVC screened OPEN/CLOSE and inserted an access method address in the DCB? Wouldn't that have allowed for easier code? :>A key question is: once the SVC Screening routine has gotten control, how :>does :>it then make sure that the "real" SVC routine gets control both in the :>right :>environment (locks included) and also with all the right data (potentially :>all :>16 64-bit GRs and ARs at the time of the SVC issuance, with the PSW of the :>issuer)? Why would that be required? Of course, it would have been nice if SVC screening did not require all intercepted SVC's to be treated as the same type. -- Binyamin Dissen <[email protected]> http://www.dissensoftware.com Director, Dissen Software, Bar & Grill - Israel Should you use the mailblocks package and expect a response from me, you should preauthorize the dissensoftware.com domain. I very rarely bother responding to challenge/response systems, especially those from irresponsible companies. ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: GET IBM-MAIN INFO Search the archives at http://bama.ua.edu/archives/ibm-main.html
