Re: Call stack inquiry performance - QWVRCSTK vs QMHSNDPM+QMHRCVPM

["Lim Hock-Chai" <Lim.Hock-Chai@xxxxxxxxxxxxxxx>]

Ok.  Use QMHSNDPM and specified the program name that I want to check on
the stack entry parameter works.  I get CPF2479 when program is not in
the stack.  (Note: Instead of *DIAG message type, I use *RQS message
type and do a QMHRCVPM to remove it.)  Not sure if this method is faster
than QWVRCSTK or not.  But it sound like it should.

Thanks Chuck.  

"Lim Hock-Chai" <Lim.Hock-Chai@xxxxxxxxxxxxxxx> wrote in message
news:<mailman.1939.1218050411.2545.rpg400-l@xxxxxxxxxxxx>...
Aahhh, sending a diag message.  Cool trick.

I'm trying to create a decrypt export procedure to allow caller to
decrypt encrypted credit card field.  Base on user authority, this
decrypt export procedure will apply appropriate masking when decrypting
the credit card field.  I'm trying to add some functionality to this
procedure where it can be setup to apply different masking base on
UserID+Caller Program Name.  Obviously, I can get the caller program
name without needing the caller passing the program name to me.  But
some how (don't know I'll ever need it yet), I have a feeling that may
be under some special condition the caller might need to override the
program to a program name that is in the higher call stack.

As far as non-RPG, well, this service program is a RPG program.  RPG has
a lot of special ability, may be and just may be, there is a special
solution that can only be found in RPG :).  I guess David will let me
know if I need to move this.

thanks



"CRPence" <CRPbottle@xxxxxxxxx> wrote in message
news:<mailman.1923.1218049031.2545.rpg400-l@xxxxxxxxxxxx>...
>    If considering QMHSNDPM to find a program in the stack, it would be

> daft to try to find entries "one level at a time."  Just send a 
> diagnostic message to the named program, and if the error CPF24xx 
> "program &x not in stack" is the result, then you know it is not in
the 
> stack.
>
>    However whenever a design exists where a program thinks it must
know 
> who called it, I start to wonder if I might not have a design problem.

> There should be very few required cases for that feature.  It might be

> worth discussing for what purpose the program name is being passed and

> why the called program finds it necessary to validate the program name

> is on the stack.
>
>    Why the non-RPG topics finding their way here of late, instead of
to 
> Midrange-L?
>
> Regards, Chuck
>
> Lim Hock-Chai wrote:
> > I've a service program that accepts program-name as one of the input
> > parameters.  When this parameter is passed in by the caller, it will
> > need to determine if the passed in Program Name is actually in the
> > current call stack.  There are two ways that I know of that can
> > determine if a program is on the call stack, one is using the
QWVRCSTK
> > api to get the current call stack;  Another way is use the
> > QMHSNDPM+QMHRCVPM to get the call stack one level at a time.  Does
> > anybody know which will perform better, or is there another way to
> > determine if a program is on the current call stack?
----------