[Info-vax] VMS Software Q1 '23 Update

[Arne Vajhøj]

On 2/7/2023 11:56 AM, Single Stage to Orbit wrote:
> On Tue, 2023-02-07 at 14:02 +0000, Simon Clubley wrote:
>>>> You can have application level bugchecks that take down the
>>>> application only and in a controlled manner. That's also what I
>>>> am talking about here in addition to the system-level bugchecks.
>>>
>>> I wasn't aware that was possible, one learns something new!
>>
>> Sorry, that didn't come across correctly.
>>
>> I meant the concept of an application shutting itself down in a way
>> comparable to a VMS bugcheck if it detects an internal integrity
>> failure.
> 
> Oh yes, that's perfectly fine.
> 
>> However, you can indeed have non-fatal bugchecks on VMS that kill the
>> process, but not the system. RMS uses them and they are (usually)
>> non-fatal to the system itself.
> 
> Makes sense.

Bug checks are not limited to E and K.

The mechanism can also be used in U and S.

To quote IDS:

"The OpenVMS operating system itself generates few bugchecks from user
or supervisor mode. It provides the mechanism for use by other
software."

I have no idea why anyone would want to use that mechanism over
more common ways to report a problem and exit.

But it is there.

On Alpha:

$ type toast.mar
         .title  toast
         .psect  $code quad,pic,con,lcl,shr,exe,nowrt
         .entry  toast,^m<>
         bug_check operator,fatal
         .end
$ macro toast + sys$library:lib/lib
$ type test.for
       program test
       call toast
       end
$ for test
$ link test + toast + sys$library:vms$volatile_private_interfaces/lib
$ run test
%SYSTEM-F-BUGCHECK, internal consistency failure

Arne