Also look at the MESSAGES dataset, for example, there is a BIG
difference between these allocations:
IGD103I SMS ALLOCATED TO DDNAME SMPLOGA
IGD100I VIO ALLOCATED TO DDNAME SYSUT1 DATACLAS ( )
IGD101I SMS ALLOCATED TO DDNAME (IMAGCOP1)
DSN (xxx.xxxx.G0013V00 )
STORCLAS (TMM) MGMTCLAS (TMM) DATACLAS (TMM)
VOL SER NOS= xxxxx2
Check the perm dataset vs. the &&temp dataset. I seem to remember a
rash of OC4's when we went to SMS, about at IBM-MAIN's half life I
think....
>-------------------------------
So maybe it is like this:
- The program always reads one more time after EOF.
- Sometimes the position after EOF is on the end of the block,
sometimes not.
- When allocated on physical disk, this doesn't matter.
- But when allocated as a &&-dataset, it is somehow (?) allocated in
virtual memory.
- And when the extra read is done when positioned at the end of the
block,
it "translates" to a read at an invalid memory address.
----------------------------------------------------------------------
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
