🚀 go-pugleaf

In article <CpJKJz.KBD@news.cis.umn.edu>, ianhogg@milli.cs.umn.edu (Ian J Hogg) writes:
> In article <JASON.94May6154541@deneb.cygnus.com> jason@cygnus.com (Jason Merrill) writes:
> >>>>>> Ian J Hogg <ianhogg@milli.cs.umn.edu> writes:
> >
> >> ...  I know of at least one compiler that passes arguments for C++
> >> linkage functions differently than C linkage functions. ...

> >...  It does seem strange to me that they would alter the calling
> >convention ...
 
> ... Zortech ... claimed [it] was more efficient.  

The 8086 processor family has two different ways to do a return from a
function.  If the _called_ function knows how many argument bytes were
pushed on the stack, it can add an operand to the return instruction; the
return will adjust the stack pointer back over the arguments while the
processor is locating and fetching the next instruction to execute after
the return.  This is sometimes called the Pascal Calling Sequence.

If the _called_ function is unsure of how many argument bytes were passed,
it cannot do this and the _calling_ function must include code to adjust
the stack pointer.  This takes extra time, and it occurs once for every call
of the function instead of once for every definition.  This is sometimes
called the C Calling Sequence.

For C, because not every function must be prototyped, the compiler cannot
assume that the correct number of argument bytes were pushed on the stack.
The faster Pascal Calling Sequence cannot be used.

For C++, because every function must be prototyped, the Pascal Calling
Sequence _can_ be used _except_ for functions whose signature includes the
ellipsis.
-- 
 (This man's opinions are his own.)
 From mole-end				Mark Terribile
 mat@mole-end.matawan.nj.us, Somewhere in Matawan, NJ
	(Training and consulting in C, C++, UNIX, etc.)