Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!mailrus!csd4.milw.wisc.edu!uxc!deimos.cis.ksu.edu!rutgers!att!mcdchg!mcdphx!teroach.UUCP!dbk
From: dbk@teroach.UUCP (Dave Kinzer)
Newsgroups: comp.sys.amiga.tech
Subject: Re: A68k and Manx AS
Summary: More 68K assembly
Keywords: 68000 assembly
Message-ID: <10891@mcdphx.phx.mcd.mot.com>
Date: 14 May 89 06:38:26 GMT
References: <221@mindlink.UUCP> <262@xdos.UUCP> <8979@polya.Stanford.EDU> <87@estinc.UUCP>
Sender: listen@mcdphx.phx.mcd.mot.com
Reply-To: dbk@teroach.UUCP (Dave Kinzer)
Followup-To: comp.sys.amiga.tech
Distribution: na
Organization: Motorola Microcomputer Division, Tempe, Az.
Lines: 28

In article <87@estinc.UUCP> fnf@estinc.UUCP (Fred Fish) writes:
>In article <8979@polya.Stanford.EDU> rokicki@polya.Stanford.EDU (Tomas G. Rokicki) writes:
>>
   The subject is optimizing 68000 branch instructions for forward branches.

   Tom explains a possible technique.

   Fred suggests assuming short branches then converting to long as needed
since once they are converted to long they will not need to be checked
again.


   I'm afraid that this falls into the pick your poison catagory.  If you
assume all branches to be long, once you start going back through the
code to shorten the branches, once they have been shortened they need
not be looked at again.  They will never fall back into the domain of
long branches (i.e. we only shrinking the code.)  It's the same thing,
only backwards.

   Assuming long first has at least one advantage:  You need not run
the optimizer if you don't want to take the time (probably minimal, but
at least you have a choice.) 



|    // GOATS - Gladly Offering All Their Support  Dave Kinzer (602)897-3085|
|   //  >> In Hell you need 4Mb to Multitask!  <<  uunet!mcdphx!teroach!dbk |
| \X/   #define policy_maker(name) (name->salary > 3 * dave.salary)         |