IBM CMS dumpfile idiocy

jim s jwsmail at jwsss.com
Wed Dec 16 15:35:08 CST 2015


 From the IBM 360/40 functional characteristics:

When IPL is initiated, the selected input device starts reading. The 
first 24
bytes read are placed in storage locations 0-23.  Storage protection, 
program
controlled interruption, and a possible incorrect length indication are
ignored.  The double word read into location 8 is used as the channel 
command
word (ccw) for a subsequent I/O operation.

When chaining is specified in this ccw, the operation proceeds with the 
ccw in
location 16.  Either command chaining or data chain- ing may be specified.

After the input operation is performed, the I/O ad- dress is stored in 
bits 21-31
of the first word in storage.  Bi~s 16-20 are made zero.  Bits 0-15 remain
unchanged.  The CPU subsequently fetches the double word in location 0 
as a new
psw and proceeds under control of the new psw.  The load light is turned 
off.
When the I/O operations and psw loading are not completed sat- isfactorily,
the CPU idles, and the load light remains on.

I think that the 24 bytes besides covering useful locations in the 
bottom of memory may be the minimum record size for the half inch drives.

So you may see 24 bytes or you may see 80.  I've been used to seeing a 
two "card" boot somewhere, but that isn't the discussion here, since we 
are talking probably a cms file.  Also may be off in the weeds, as far 
as this is concerned.

if you bring up a copy of the VM/370, and convert this to an AWS tape, 
on some available CMS drive, you should be able to read it in with some 
command line incantation.  How you would edit it from there with the 
public domain VM/370's editing is anyones guess. XEDIT is licensed, so 
you won't have something easy to use to edit and save it if it is a text 
file.

By the time you figure all that out, assuming you don't regularly use 
either VM/370 or MVS 3.8 on Hercules, or have access to IBM facilities 
to use actual IBM products, coding and squinting at your results will 
probably produce the result faster.  I'm guessing you already have it in 
process anyway.

thanks
JIm

On 12/16/2015 12:35 PM, Chuck Guzis wrote:
> On 12/16/2015 12:03 PM, jwsmobile wrote:
>
>> Does the first block have a PSW looking thing in the appropriate
>> start, and an I/O program to boot the next record from the boot
>> device?
>
> Not to my eye--it starts off thusly:
>
> 02 c3 d4 e2 c6
>
> Looks more like 02 "CMSF"
>
> --Chuck
>
>



More information about the cctalk mailing list