Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!uwm.edu!uakari.primate.wisc.edu!aplcen!ginosko!uunet!ibmpa!ghoti!lmb
From: lmb@ghoti.uucp (Larry Breed)
Newsgroups: comp.lang.c
Subject: Re: swap(x,y)
Message-ID: <2155@ibmpa.UUCP>
Date: 16 Sep 89 07:59:09 GMT
References: <MYwDEFy00WB842JEUx@andrew.cmu.edu> <8350@boring.cwi.nl> <10897@smoke.BRL.MIL> <1547@l.cc.purdue.edu>
Sender: news@ibmpa.UUCP
Reply-To: lmb@ibmsupt.UUCP (Larry Breed)
Organization: IBM AWD Palo Alto
Lines: 22

A good compiler will produce better code for the naive algorithm than
can be produced for any xor technique.  Here's the code generated by
the Metaware compiler on AOS (IBM's 4.3 for the rt pc).  

#void swap(int* i, int* j)
_.swap:
#{
#   int t;
#   t=*i;
        ls      r0,0(r3)
#   *i=*j;
        ls      r4,0(r2)
        sts     r0,0(r2)
#   *j=t;
        brx     r15		# branch with execute of following instruction
#}
        sts     r4,0(r3)


Disclaimer: Don't blame my employer, blame:
Larry Breed			(415) 855-4460
uucp: uunet!ibmsupt!lmb		inet: ibmsupt!lmb@uunet.uu.net