On 2015-08-20, at 05:39, Peter Relson wrote:
> 
>> Does this mean that one can't create a data module to be loaded 
> above-the-bar?
>> One must obtain STORAGE and copy into it?  Can one IDENTIFY an address 
> above-
>> the-bar?
> 
> No to all three.
> You can create a data module to be loaded above the bar. You can use LOAD 
> with ADDR64 to get the system to load into the area that you provide.
>  
A disappointing restriction.  I hope full Content Supervision support
extends above the bar someday.  (Is it constrained by existing Data
Areas definitions?)

> Nobody would implement an option that was clearly supporting incorrect 
> information whose sole intent is to avoid some processing you don't think 
> you want.
>  
Technically, one might argue that any AMODE value is "incorrect" for a
pure data object.  The programmer ought to be allowed to choose the
most useful value.

> Now if you had mentioned some new option asking for a non-AMODE-identified 
> return value that could be different.
>  
Such an option might be useful if a given ENTRY point could be used
alike for branch and non-branch operations

> If you don't need bit 63 because you're not doing a BASSM then clear it 
> (whether because it's data or because you're doing a BASR or whatever). 
> And simply put: don't place AMODE 64 data that is to be externally located 
> on an odd address unless you know that it will be on an odd address (so 
> that the user would know to avoid clearing bit 63).
>  
"know" requires the programmer to make undesired assumptions about the
address of an object.

How are bit 32 and bit 63 handled in 64-bit V-CONs?  (Do 64-bit V-CONs
exist?)

-- gil

Reply via email to