Path: utzoo!utgpu!bnr-vpa!bnr-rsc!tak
From: tak@bnr-rsc.UUCP (Mike Takefman)
Newsgroups: comp.arch
Subject: Re: taken -vs- untaken branches, Fortran FREQUENCY declaration
Summary: What about intrusiveness?
Message-ID: <604@bnr-rsc.UUCP>
Date: 24 Jan 88 21:04:55 GMT
References: <839@ima.ISC.COM> <2158@geac.UUCP>
Organization: Bell-Northern Research, Ottawa, Canada
Lines: 23

In article <2158@geac.UUCP> john@geac.UUCP (John Henshaw) writes:
.	1. Design, code and test your program.
.	2. Compile it for "branch profiling".
.	3. Execute the program on a dataset large enough that there is
.	   sufficient confidence that the program's execution is truly
.	   representative of "standard production activity".
.	4. Recompile the program. This recompilation should use the
.	   information gained from step 3.
.There are a few assumptions here :-). I think they're obvious.

One assumption that may not be obvious but is important is that this 
method is highly intrusive. This method will not work well for a 
real time application. 

For vanilla programming I fully endorse this view, but I believe 
that the challenge is in doing this non-intrusively.


-- 
Michael Takefman	       There are 2 basic truths,          
Processor Exploratory	         Never fight a land war in Asia and
tak@bnr-rsc@bnr-vpa@utzoo....      never go against a Sicilian if death is 
Bell Northern Research               on the line.   (The Princess Bride)