The classic dirty trick I saw was a branch 'ladder' where R15 had a return
code (must be in multiples of 4!) and the first branch would jump to one of
several following branches to real code.

Woe betide if R15 held 3, -1 etc :-)

Roops



On Sat, 2 Aug 2025 at 22:28, Charles Mills <[email protected]> wrote:

> > Did that gain the programmer any advantage over such as TM; BC or was it
> entirely theatrical?
>
> I don't recall at all.
>
> You can certainly picture some sort of elaborate branch table type logic
> dependent on both a condition and an index register in the EX register.
>
> Charles
>
> -----Original Message-----
> From: IBM Mainframe Assembler List <[email protected]> On
> Behalf Of Paul Gilmartin
> Sent: Saturday, August 2, 2025 1:27 PM
> To: [email protected]
> Subject: Re: Execute-Type Instructions
>
> On 8/2/25 07:54, Charles Mills wrote:
> > I think I recall code back in the day that did clever "computed branch"
> logic by making a BC or BCR instruction the target of EX.
> >
> Did that gain the programmer any advantage over such as TM; BC or was it
> entirely theatrical?
>
> > I would defenestrate any programmer that did that sort of thing today.
> >
> Is the ALTER instruction alive and well in COBOL?
>
> --
> gil
>

Reply via email to