Segment Translation error

Julian Wall

unread,

Jul 30, 2012, 10:37:48 AM7/30/12

to

I have a DOS RPGII program that has 1 input file (READ01) a Input demand file (KSDS) and a print file that prints a banner for report distribution. Occasionally it a abends with a segment translation error, always at night too, Usually I can restart it and it runs the next time. Looks like its in 31 bit mode and in a $$BCVS02 module. I am wondering if its a getvis problem, its in a dynamic partion that vsize=2048k and getvis is 896k. or could it be a system getvis problem.

V1 0046 // JOB MS13A PAYLESS OPEN ORDER LIST MS0032 00:00:50

DATE 07/30/2012, CLOCK 00/00/50 00:00:50

V1 0046 00:00:50

V1 0046 * JOB BT0001 00:00:50

V1 0046 0S03I PROGRAM CHECK INTERRUPTION - HEX LOCATION 000082BC - 00:00:51

INTERRUPTION CODE 10 - SEGMENT-TRANSLATION EXCEPTION 00:00:51

V1 0046 0S08I LOG. TRANS. AREA CANCELED, PHASE = $$BCVS02 00:00:51

*** MESSAGE REDISPLAY *** 07/29/12 23:07:34

0S03I PROGRAM CHECK INTERRUPTION - HEX LOCATION 000082BC - INTERRUPTION CODE 10

0S08I LOG. TRANS. AREA CANCELED, PHASE = $$BCVS02

0S00I JOB MS13A CANCELED

0S07I LOG. TRANS. AREA PSW = 070D1000 800082BC

Martin Trübner

unread,

Jul 30, 2012, 11:17:19 AM7/30/12

to

Julian,

>> I am wondering if its a getvis problem, its in a dynamic partion that
vsize=2048k and getvis is 896k. or could it be a system getvis problem.

First of all you need at least a partition dump.

If the problem is in some code get the dump to the owner of the code
and let them analyse it (and fix as well)

Second: Yes a problem with GETVIS could manifest in such a way- But
again: the code should not die but handle it - and if it does not
want to handle it, let the first failure occur (which is usaly
documented).

I give you a sample: I am lazy and do not want to handle problems
which rarely occur. For this very reason I have changed all my GETVIS
request to STORAGE OBTAIN (which has the option to tell the op-sys to
kill if it can not be satisfied) - Much better than any unique error
message

This can be done with almost everything else.

You could of course find a circumvention- but I do like that approach
only if everything else failed.

--
Martin

Pi_cap_CPU - all you ever need around MWLC/SCRT/CMT in z/VSE
more at http://www.picapcpu.de

_______________________________________________
VSE-L mailing list
VS...@lists.lehigh.edu
https://lists.lehigh.edu/mailman/listinfo/vse-l

mick poil

unread,

Jul 30, 2012, 3:24:36 PM7/30/12

to

Do I assume that you can't open a PMR? Unless somebody else has the answer, it is normally simple for cases like this - it is called TPE.

The Technician Proximity Effect causes software to work when otherwise it won't. That is, when you stand by the d*mn machine (PCs in particular) it works perfectly. Give up your life and stay at the office and it will always work.

:-)

Alternatively, get a // OPTION DUMP and try and debug the VSE transient.

(P.S. A PMR is best. Even I don't have access to the VSE code so I can't help at that level).

(P.P.S hope you saw the smiley)

Julian Wall

unread,

Jul 31, 2012, 9:44:54 AM7/31/12

to

Thanks all, Ill try to get our sys prog to see if he wont get a PMR..

Julian

mick poil wrote: