Re: Permissions of IFS file created with QZRUCLSP

[Barbara Morris <bmorris@xxxxxxxxxx>]

On 3/18/2011 9:38 AM, Dennis wrote:
> Ah, that sheds some light!  Since we have no (apparent) means of
> telling QZRUCLSP that certain parameters are optional but passed, it
> seems unlikely that it would be capable of guessing the correct means
> of passing same.
>
> Probably a moot point, but is the "differently" part of your message
> a trade/business/compiler secret?

It's more that I don't know all the details since it's mostly 
transparent to the compilers.  The compilers just tell the optimizating 
translator that the called function has optional parameters.

I do know that for a "..." C function, the function would use the 
va_list mechanism to access optional parameters.  This requires that the 
optional parameters be contiguous in storage, so parameter 3 follows 
parameter 2 in storage, at a fixed offset away.  When there are no 
optional parameters, the system can pass the parameters in whatever way 
makes the most sense.  So the location of parameter 3 can't be predicted 
from the location of parameter 2.