From: Paul Shirley <Paul AT chocolat DOT obvious DOT fake DOT foobar DOT co DOT uk>
Newsgroups: comp.os.msdos.djgpp
Subject: Re: Infinite loop???
Date: Fri, 10 Jul 1998 19:17:28 +0100
Organization: wot? me?
Message-ID: <VqrTZAA4qlp1Ew16@foobar.co.uk>
References: <35A3D8C3 DOT 208E6D23 AT cyberdude DOT com> <35A5FC72 DOT C235F6D9 AT eik DOT bme DOT hu>
NNTP-Posting-Host: chocolat.foobar.co.uk
Mime-Version: 1.0
Content-Type: text/plain; charset=iso-8859-1
Content-Transfer-Encoding: 8bit
Lines: 24
To: djgpp AT delorie DOT com
DJ-Gateway: from newsgroup comp.os.msdos.djgpp
Precedence: bulk

In article <35A5FC72 DOT C235F6D9 AT eik DOT bme DOT hu>, Dr. András Sólyom
<solyom AT eik DOT bme DOT hu> writes
>For the left shift operator BCC uses an 
>       SHL (shift left)
>instruction, while GCC uses 
>       SAL (arithmetic shift left)
>I think these differ in how they uses the CARRY bit
>
>why is it so? 
>       AFAIK the implementation of the left shift operator is left to the 
>compiler
>maker by the ANSI C specification.

The >> operator is clearly defined. For signed operands it does an
arithmetic shift, unsigned ops get a zero extending shift. If you see
anything different its a compiler bug.

You should check the type of the value being shifted to explain the
difference you see. In particular shifting a char type may cause
differences between compilers because a char may be signed or unsigned
by default in older compilers.

---
Paul Shirley: my email address is 'obvious'ly anti-spammed