Tom,
Thanks for you reply. Makes sense.
--Dave Day
----- Original Message -----
From: "Tom Schmidt" <[EMAIL PROTECTED]>
Newsgroups: bit.listserv.ibm-main
To: <IBM-MAIN@BAMA.UA.EDU>
Sent: Thursday, March 22, 2007 11:27 AM
Subject: Re: VSAM/DEB/ACB question
On Wed, 21 Mar 2007 18:27:27 -0600, David Day wrote:
Have not done much VSAM coding, hence the questions. Have an opened
ACB in 31 bit storage. Looking at a dump immediately after a successful
OPEN
for the dataset. TCBDEB points to a DEB where the chain from the DEB to
the
JFCB is correct(DEB extention to DSAB to SIOT to JFCB)...it's the dataset
just
opened. At x'14' into the DEBBASIC is a pointer to the ACB, but the field
in
the DEB is called DEBECBAD. The DEBDCBB field is filled in, but don't
recognize what it points to. Are the differences between VSAM and non-VSAM
DEB structures and associated control blocks documented somewhere?
The field in the DEB that points to an ACB (or DCB) is called DEBDCBAD
(not
DEBECBAD -- that's the pointer to the plist used to locate the purge ECB
for
an SVC purge request). And you don't REALLY want to reference the 4-byte
DEBDCBAD field -- you should be using the 3-byte DEBDCBB field instead.
The high-order byte of DCBDCBAD will often be x'8F' (the 8 represents the
task
protection key and the F is a constant to identify the control block as a
DEB).
With that relatively constant x'8F' in the high-order byte of the word you
have
a tough time mapping above the line, which is where the 24-bit restriction
originates.
DCBs cannot be located outside of 24-bit virtual storage. I don't believe
that
ACBs can be located outside of 24-bit virtual storage either, but there is
a
caveat to this: there are 2 kinds of ACBs (and DCBs) -- Protected and
Unprotected. The user's ACB (or DCB) is the Unprotected flavor. The
system
produces a copy in protected storage (hence the term) which is the actual
xCB used by Open/Close/etc. The status of the Protected xCB is reflected
back to the user's Unprotected xCB as the I/O operations progress.
This is documented (more or less) in the various DFSMS (DFP) publications
...
and a whole lot of older, not-completely-obsolete PLMs and FE handbooks.
--
Tom Schmidt
Madison, WI
----------------------------------------------------------------------
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
----------------------------------------------------------------------
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
