Path: utzoo!utgpu!jarvis.csri.toronto.edu!cs.utexas.edu!usc!snorkelwacker!spdcc!mirror!ima!esegue!compilers-sender
From: Moss@cs.umass.edu (Eliot Moss)
Newsgroups: comp.compilers
Subject: Re: Register Allocation
Message-ID: <1990Feb26.214750.10649@esegue.segue.boston.ma.us>
Date: 26 Feb 90 21:47:50 GMT
References: <1990Feb20.155619.3121@esegue.segue.boston.ma.us>
Sender: compilers-sender@esegue.segue.boston.ma.us
Reply-To: Moss@cs.umass.edu (Eliot Moss)
Organization: Dept of Comp and Info Sci, Univ of Mass (Amherst)
Lines: 28
Approved: compilers@esegue.segue.boston.ma.us
In-Reply-To: napi@rangkom.MY's message of 20 Feb 90 15:56:19 GMT

In article <1990Feb20.155619.3121@esegue.segue.boston.ma.us> napi@rangkom.MY (Mohd Hanafiah b. Abdullah) writes:

  How does one perform register allocation on scalar variables that may
  be pointed to by others?

  I am curious because, a pointer assumes that the variable it points to
  resides in memory, but in actuality the variable resides in a register.

Well, you either do not allocate it to a register, or you propagate its value
to its memory home whenever (a) it may have changed since the last time you
did so, and (b) there is a possibility that the memory value may be picked up
through an aliasing pointer. If (a) and (b) occur together often enough, then
there is little benefit to putting the variable in a register, but if they
occur rarely compared with use of the variables, then there may be benefit. Of
course, you could use a machine such as the DEC PDP-10 where the registers
*are* addressable memory (the first 16 locations) ... :-)      Eliot
--

		J. Eliot B. Moss, Assistant Professor
		Department of Computer and Information Science
		Lederle Graduate Research Center
		University of Massachusetts
		Amherst, MA  01003
		(413) 545-4206; Moss@cs.umass.edu
-- 
Send compilers articles to compilers@esegue.segue.boston.ma.us
{spdcc | ima | lotus}!esegue.  Meta-mail to compilers-request@esegue.
Please send responses to the author of the message, not the poster.