Subject: Re: controlling where go32 puts program
To: eliz AT is DOT elta DOT co DOT il
Date: Tue, 12 Jul 1994 09:42:36 -0600 (CDT)
Cc: djgpp AT sun DOT soe DOT clarkson DOT edu
From: mcastle AT umr DOT edu (Mike Castle)

Amazingly enough eliz AT is DOT elta DOT co DOT il said:
> I think, some of GO32 has to stay in memory also because of the extender code
> required to provide DOS services to 32-bit programs (such as gcc itself).  So
> it's not only the argument-passing thing.

Ahh.. but once that 32-bit program calls system(), then the go32
support is no longer needed, and could be swapped out.  

Though I just thought of a new concern, inspired by DJ's response
to QDPMI question.  

What about chaining interrupts?  Let's say go32 was swapped out,
then the mouse was moved, what happens?  Is that mouse handling
code swapped out?  If so, does some bogus info gets called?

For example, suppose someone wrote a go32 based mouse driven
menu.  If the child program doesn't use the mouse, and the mouse
is moved, isn't an interrupt still made, and the handler needed
to be called?  Would this require having to fix up vectors and
and unfixing them upon return?  (Then again... is this what the
borland routines swapvector() do, and why borland recommends you
do something like swap();system();swap() ??)

mrc
-- 
Mike Castle .-=NEXUS=-.  Life is like a clock:  You can work constantly
  mcastle AT cs DOT umr DOT edu     and be right all the time, or not work at all
   mcastle AT umr DOT edu       and be right at least twice a day.  -- mrc
    We are all of us living in the shadow of Manhattan.  -- Watchmen