I agree fully that there is nothing to worry about. However, for the existing program I do not want to bother with ARCHLVL setting as it requires me to analyze all code changes resulting from ARCHLVL=2.
However, for new programs I will take over your suggestion in the previous post. -- Thanks a lot, Manfred On Fri, Sep 27, 2013 at 6:11 PM, Ed Jaffe <[email protected]>wrote: > On 9/27/2013 1:21 AM, Manfred Lotz wrote: > >> The change is this: >> > > [snip] > > There is nothing to worry about. The two expansions make the identical > service call (as seen by the operating system), but the technique used by > the updated expansion is usable by programs that use the relative & > immediate instruction facility (you get that with ARCHLVL=1 or higher). > > Basically, the inline parameters and the 'L 15,IHB0004F' instruction in > the old-style expansion would force a modern program to establish temporary > code base register coverage just for the GETMAIN macro, which is ugly code > that uses an extra register and slows down the program a little bit. The > newer expansions remove that restriction by moving the inline parameters > into the literal pool and can be safely by older programs. > > > -- > Edward E Jaffe > Phoenix Software International, Inc > 831 Parkview Drive North > El Segundo, CA 90245 > http://www.phoenixsoftware.**com/ <http://www.phoenixsoftware.com/> > > ------------------------------**------------------------------**---------- > For IBM-MAIN subscribe / signoff / archive access instructions, > send email to [email protected] with the message: INFO IBM-MAIN > ---------------------------------------------------------------------- For IBM-MAIN subscribe / signoff / archive access instructions, send email to [email protected] with the message: INFO IBM-MAIN
