> I'm not sure that any of the hardware architecture has been implemented 
> specifically for Linux.

I certainly do not know one way or the other whether any hardware features were 
implemented specifically for Linux, but I do not find it impossible, given that 
there is a whole series of models marketed specifically for Linux (the LinuxONE 
series, of course).

Charles


-----Original Message-----
From: IBM Mainframe Discussion List [mailto:[email protected]] On Behalf 
Of Tony Harminc
Sent: Thursday, June 20, 2019 11:59 AM
To: [email protected]
Subject: Re: Use of "trap" facility in z/OS?

On Thu, 20 Jun 2019 at 10:58, Farley, Peter x23353
<[email protected]> wrote:
>
> Some time back I started a thread on this forum about writing authorized code 
> or a PC routine to update the DUCT
> architectural control block to populate the TRAP fields to make active use of 
> the variety of TRAP instructions now provided,
> and the common consensus at the time was for me to keep my grubby hands off 
> the DUCT lest I cause a catastrophic and
> possibly ipl-able failure.
>
> IIRC there was a comment in that thread by one of our regular IBM 
> contributors that at least one IBM product had gone ahead
> and used that technique without dispensation from the core z/OS developers 
> who were (at least at the time) not happy about
> the choice made by those product developers.

The interesting question, in my opinion, is how such a scheme got to
be in the architecture without z/OS software infrastructure
to use it. Dan Greiner commented in another thread about how BIC came
to be, and I understand that in recent years there has been close
collaboration between the hardware architects and the software people
in core z/OS, compilers, LE, and particularly Java groups. But if
there's no z/OS support...

> We have yet to see any further word from IBM on when or if z/OS will provide 
> the needed system-level interfaces to "safely"
> use TRAP and friends.  "Business justification needed" and "ROI" are of 
> course the usual reasons given for delaying support
> for the architected hardware capabilities.  Hiring more core z/OS developers 
> and testers to keep up with support for all of the
> hardware capabilities never seems to be on the list of possible solutions.
>
> z/Linux may or may not be ahead of z/OS in regard to actually using TRAP and 
> friends, but I don't know where one would look
> to find out if and/or how it is implemented there.

I'm not sure that any of the hardware architecture has been
implemented specifically for Linux. Certainly the IBM Linux coders
have shown themselves very adept at exploiting existing features in
ways that were perhaps not intended or anticipated by the hardware
architects.

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

Reply via email to