RE: Fighting cfint (academic quw -- MI400

Far be it for someone as unknowledgeable on the subject as I to burst anyone's
bubble on this, but it seems to be folly to think that IBM would have done
anything less than locked the CFINT animal down tight, hidden in the deep,
unknown recesses of the OS.  Even if you could find "it", what would you do
with it?  Mess with it, and you render your box useless?  Or maybe it sends a
wireless message to IBM Rochester?

Of course, IBM wanted everyone to think that AS/400 password security was
ironclad, too, didn't they?  Well, then, have at it!  And be sure to post your
results!  <g>

Dan Bale
IT - AS/400
Handleman Company
248-362-4400  Ext. 4952

-----Original Message-----

In theory one would think you are correct. Since CFINT01 is implemented in
code one should be able to get around it. However, CFINT01 is a strange
task. It shows up in WRKSYSACT as a task, but I can't find a TDE for it in
SST. I can find other tasks, but not CFINT01. I'd bet IBM changed SST to
prevent CFINT from showing up.

Does anyone know how to access the Task Dispatch List itself? I can see
entries from SST, but I want to see the hex in the TDL itself. CFINT01
_must_ be in there for it to run, but I can't find the TDL. Any ideas? An
address would be great? Or an offset in the MISR or NACA

-Walden

-----Original Message-----
From: shahar mor [mailto:shahar_mor@yahoo.com]
Sent: Tuesday, October 17, 2000 4:00 PM
To: MI400@midrange.com
Subject: Fighting cfint (academic quwstion)

Hi all,

This is purely academic question yet i'm aruing this
issue with a collegue for sometime now.
so i'm asking out of 'academic curiosity'

The facts:

ibm mistreats it's installed base. Whenever you buy a
'server model' and use interactive application a
looped task named cfint starts eating your cpu.

i'm sure this task is needed only to make current
customer pay more for the existing applications and it
is not hardware related (here the argue begins..)

If i'm correct and this cfint is software based then
in teory somehow one may expect to 'cheat it' if it is
hardware then even in teory this is not doable.

Bottom line:

Am i right ? if i had the internal knowledge will it
be possible to slow down "crazy" cfint thus paying
less for the same performance ?

Thanks for any input, and once again i'm NOT trying
anything harmfull in this post(although ibm deserves
criticism over their current policy).

Regards

=====
Shahar mor
consultant
+---
| This is the MI Programmers Mailing List!
| To submit a new message, send your mail to MI400@midrange.com.
| To subscribe to this list send email to MI400-SUB@midrange.com.
| To unsubscribe from this list send email to MI400-UNSUB@midrange.com.
| Questions should be directed to the list owner/operator: dr2@cssas400.com
+---