On Sat, 22 Jul 2006, George Birbilis wrote:

> > > It will only work for Intel 32-bit since it uses assembler
> > > instructions at some parts. BTW, I though FPC assembler isn't
> > > compatible with Delphi one,
> >
> > That's a wrong though ;) Almost all inline assembler written
> > for delphi works with fpc.
> 
> I'm sure I read that thing the other day:
> I think it's at http://www.freepascal.org/port.html, but maybe it meant the
> Turbo Pascal inline assembler statemets, not the Delphi ones? (I thought
> they were same or similar in Delphi, don't remember)
> 
> > > maybe Boguslaw fixed that code?
> > > Best thing would be to have compiler defines and rewrite those
> > > assembler instructions as normal Pascal code (would help porting to
> > > other CPUs)
> >
> > This is probably impossible, PS does a lot of dirty tricks
> > with subroutine calls etc.
> 
> I think Carlo said sometime ago on the ROPS newsgroup that assembler is used
> mostly for maintaining the call stack. In fact I don't think anything done
> using assembler can't be done with normal Pascal code (if necessery it can
> call into a unit of low-level primitives that will be implemend in
> assembler, but those primitives would be easily portable among various CPU
> architectures)

That is exactly what it does: there are 1 or 2 assembler routines, to create a
correct call stack. You can't do this in Pascal. These 2 routines would be the
only thing you need to port accross CPUs...

Michael.

_________________________________________________________________
     To unsubscribe: mail [EMAIL PROTECTED] with
                "unsubscribe" as the Subject
   archives at http://www.lazarus.freepascal.org/mailarchives

Reply via email to