Path: utzoo!utgpu!water!watmath!clyde!att!ucbvax!decwrl!sun!limes@ouroborous
From: limes@ouroborous (Greg Limes)
Newsgroups: comp.lang.c
Subject: Re: C IF statement efficiency...
Keywords: c if
Message-ID: <65327@sun.uucp>
Date: 23 Aug 88 17:41:34 GMT
References: <8808171400.AA05122@ucbvax.Berkeley.EDU> <566@sybase.sybase.com>
Sender: news@sun.uucp
Reply-To: limes@ouroborous (Greg Limes)
Distribution: na
Organization: Sun Microsystems, Inc.
Lines: 20
In-reply-to: ben@bosco.UUCP (ben ullrich)

In article <8808171400.AA05122@ucbvax.Berkeley.EDU> U23405@UICVM writes:
>I was wondering if, of the two following program fragments, which one would
>be compiled more efficiently by most C compilers:
>
>          if (big > small)             if (big > small)
>               return big;                  return big;
>          else                         return small;
>               return small;

In article <566@sybase.sybase.com>, ben@bosco (ben ullrich) writes:
>how about
>	return ((big > small) ? big : small)

>i don't know how efficient this is, but it is certainly more succinct than
>those suggested in the quoted text above.

All three of these examples reduce to the same thing quite early on in the
compilation; in fact, the compiler on my workstation (SunOS 4.0) produces code
for them that differs only in the names of some of the temporary labels used.
--