To reinforce Tony's point: ultimate control resides with SAF update authority 
to any and all authorized libraries. If that control is compromised, there is 
NOTHING that MVS can do to prevent mischief. 

.
.
J.O.Skip Robinson
Southern California Edison Company
Electric Dragon Team Paddler 
SHARE MVS Program Co-Manager
323-715-0595 Mobile
626-543-6132 Office ⇐=== NEW
[email protected]

-----Original Message-----
From: IBM Mainframe Discussion List <[email protected]> On Behalf Of 
Tony Harminc
Sent: Friday, November 15, 2019 10:19 AM
To: [email protected]
Subject: (External):Re: AUTHPGM in IKJTSOxx

On Wed, 13 Nov 2019 at 09:56, Jeffrey Holst 
<[email protected]> wrote:
>
> Does AUTHPGM require that the specified program have a non-zero AC or that it 
> be in an APF authorized library?

Both.

> I ask because it appears that a very clever user may have written a program 
> whose name matches a program in the AUTHPGM list. The program executes a 
> macro instruction that requires APF authorization. It appears that he was 
> able to successfully call it from TSO.

If the user has write access to an APF authorized library (including any 
library in linklist), then all is lost - the user can do anything.
If the user does not have such write access, then how did s/he invoke this 
bogus program? With TSO CALL specifying a dataset name? If an unauthorized 
library is in the STEPLIB (or TSOLIB defined library or is the library 
specified on the CALL command), and the module name is defined in AUTHPGM, then 
it should get an abend S306 and/or a message from TSO saying that it can't be 
invoked.

In passing, anyone can create a load module/Program Object that has AC(1). This 
is just a mark by the creator that this module is intended to be safely invoked 
as the first module in a job step or directly as a TSO command. It bestows no 
APF authorization by itself.

> If this is the case, is there a way to secure this. If this is not supposed 
> to work this way, this would seem to be an integrity issue that is worthy of 
> a PMR.

I'm sure IBM will treat it very seriously if you can demonstrate that a user 
with *no write access to an APF authorized library* can have their own program 
- no matter what it's named -  invoked in an authorized state.

Tony H.


----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN

Reply via email to