Path: utzoo!utgpu!jarvis.csri.toronto.edu!rutgers!apple!usc!hacgate!ashtate!dbase!awd
From: awd@dbase.UUCP (Alastair Dallas)
Newsgroups: comp.lang.c
Subject: Re: iAPX86 code for ABS (using Microsoft C?)
Summary: MSC does not inline abs()
Message-ID: <121@dbase.UUCP>
Date: 20 Jun 89 23:47:11 GMT
References: <1202@draken.nada.kth.se>
Organization: Ashton Tate Devlopment Center Glendale, Calif.
Lines: 36

In article <1202@draken.nada.kth.se>, d88-eli@nada.kth.se (Erik Liljencrantz) writes:
> I remember a review of Microsoft's C compiler for MS-DOS that stated it
> produced no-conditional-jump-abs-code. The function to convert an integer
> to a positive integer was performed without a conditional jump (i.e. JS or
> something). I'm very interested in this piece of code, so if someone
> who does have a Microsoft C-compiler could try something like
>   main()
>   { int a,b;
>     a=-10;
>     b=abs(a);
>   }
> and take a close look at the assembly code generated, I would appreciate
> that. This kind of reversed engineering (or what is it?) can't be illegal
> as it actually was published in a magazine some time ago (maybe two years
> so it must have been Microsoft C version 4.00 that was reviewed...)

This code, using Microsoft C v5.1, produces:

	...
	mov	ax,FFF6
	push	ax
	call	_abs()
	...

The abs() function uses conditional jumps.  As for reverse engineering not
being illegal, I think you are probably mistaken.  If I were to broadcast
here the code for the abs() function (which is trivial) so that people
who had not bought MSC could take and use (read: steal) it, it would be
wrong.  On the other hand, if there were some trick code for doing this
on an Intel processor, and if it were in Microsoft (or any other) C,
I would bet the trick did not originate with the C vendor's compiler team.
There's a difference between research and various nastier terms.  

/alastair/

Disclaimer: I speak for myself, only.