Unconfigured Ad Widget

Collapse

Announcement

Collapse
No announcement yet.

RPGII Call error on 9402-236

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • RPGII Call error on 9402-236

    We are probably the last users on earth who still operate the S36 box 9402-236 (the Advanced 36 running SSP 7.1 ) and are looking for someone who is familiar with the RPGII "CALL" function to pass control to an seperate, external program. We get an occasional RPG I/O error (9071) from a program using several of these CALL functions and are looking for some kind of input regarding it. Any info, leads, contacts would help. Thanks, Ron Macomber, MIS The Warren Co. Erie, Pa. warrenco@erie.net 1-800-562-0357

  • #2
    RPGII Call error on 9402-236

    You are not the last. Believe it or not, there are still shops running 5360 S/36s. If memory serves, the AS/36 RPG II "CALL" command works precisely like the RPG III CALL. Either a PLIST may be specified in factor 2, or PARM(s) may be placed on subsequent lines, with the fields used as parameters specified in factor 2. Whichever method is used, the PARM(s) of the calling program must match the type and length of the PARM(s) of the called program. Dave

    Comment


    • #3
      RPGII Call error on 9402-236

      Whose compiler are you using? IBM, ASNA, or BPS.

      Comment


      • #4
        RPGII Call error on 9402-236

        IBM's. Details are as follows: System is 9402-236/SSP 7.1, IBM RPGII, PTF level E6: We have a large screen program that CALLs another small screen program, passing 1 parameter to and from the called module (parameters match). Both programs use command keys and roll keys. The called program occasionally, particularly at peak usage times, will fail giving the error "RPG-9071, I/O exception occurred to device Wx" (workstation i.d.) and the program must be ended with an option (3). When it fails, DEBUG shows that the calling program does what it should, control is passed to the called program which displays an input/output screen, and waits for input. When a command key is pressed to start another RPG cycle and return to the calling program, the RPG error occurs. DEBUG shows that the called program does not even go to 1st cycle when the command key is pressed , as though an error occurred at exception input time when the command keys are input to the program, and of course option (3) which is the only option ends the called program and does not return to main program. Any input would be appreciated. Thanks, Ron Macomber, MIS The Warren Co. Erie, Pa. warrenco@erie.net 1-800-562-0357

        Comment


        • #5
          RPGII Call error on 9402-236

          Could you post some code snippets from the calling and called programs? Dave

          Comment


          • #6
            RPGII Call error on 9402-236

            Thank you for the reply. We have checked the code and it seems correct. The program works 99% of the time just as it was intended it to do. Only occasionally does the RPG-9071 error occur, usually at peak times- just enough to cause an annoyance as the program has to be ended with an option (3) which ends it without finishing any more job steps. Most often the error is generated from a remote site using a p.c. and remote 5250 emulation over the web to a token ring p.c. at the main facility (BOS Advanced Server for SAA) although it happens with local workstations almost as often. The calling program is 4400 lines and the 5 called programs are 400-600 lines of code each. The called programs are all called individually by the base program (not chained) and sometimes 2 are called in the same cycle of the base program and sometimes only 1 is called in a cycle. The problem only occurs in the 1 called module that inputs command keys and input fields- the other called programs have either no screen i/o at all or do not have the combination of input fields and command keys. The section of the calling prog looks like this: (some input fields...) (some prior calcs...) DEBUG CALL C#PRG2 PARM FDL1 (input only) PARM FLD2 (input only) PARM FLD3 (updated and passed back) PARM FLD4 (input only) PARM FLD5 (input only) FREE C#PRG2 DEBUG (some more calcs, display results & end cycle) The section of the called program, C#PRG2 looks like this: (some input fields...) >>-------------------------------------------------------------------->> this is where the RPG i/o error sometimes occurs, DEBUG is never reached on 2nd cycle as though the error occurs at input time for KC) >>-------------------------------------------------------------------->> DEBUG N50 GOTO START (1st cycle only) *ENTRY PLIST PARM FLD1 PARM FLD2 PARM FLD3 PARM FLD4 PARM FLD5 EXCPT SCRN1 SETON 50 GOTO END START TAG DEBUG KC RETRN KC GOTO END (end of cycle, release workstation file, return to calling program) (do some calcs, display data with exception output) TAG END We programmed by the book and as I say, it usually works fine. Ron Macomber,MIS The Warren Co. Erie, Pa. warrenco@erie.net 1-800-562-0357

            Comment


            • #7
              RPGII Call error on 9402-236

              Results of further investigation: Using the RPGII INFDS data structure and the INFSR subroutine for workstation file exception/error handling and getting the value of the *STATUS code word and positions 23-26 which contains a 4-character return code, we no longer get the RPG error #9071 (i/o exception error for workstation file). The program now handles the exception error and the subroutine we coded into it lists a DEBUG printout and redisplays the last screen so there no fatal error anymore. The *STATUS value returned is "01331" (time interval expired). The value of the data structure position 23-26 is "0310", which is not listed in the RPG manual; it says to refer to the IBM manual "Interactive Communications Feature: Programming for Subsystems and Intra Subsystem Reference" , which we do not have. If anyone has this manual, I would appreciate a copy of the code chart for these values. At least the program doesn't bomb anymore, but it would be nice to know what's going on. If anyone has any ideas I would appreciate hearing them. Thanks to all for the input so far, Ron Macomber The Warren Co. Erie, Pa. warrenco@erie.net

              Comment

              Working...
              X