home *** CD-ROM | disk | FTP | other *** search
/ Executor 2.0 / executorv2.0.iso / pc / dos / extra / docs / maillist / text / archive.95 / text0637.txt < prev    next >
Encoding:
Internet Message Format  |  1996-03-31  |  1.0 KB
  1. From: Arnt Gulbrandsen <agulbra@troll.no>
  2. To: executor@nacm.com
  3. Subject: Re: 486 optmization
  4. In-Reply-To: <Pine.LNX.3.91.950427073305.2351A-100000@zeus.achilles.net>
  5. Message-Id: <Pine.LNX.3.91.950427200945.14728B-100000@popcorn.troll.no>
  6. Mime-Version: 1.0
  7. Content-Type: TEXT/PLAIN; charset=US-ASCII
  8. Sender: owner-executor@nacm.com
  9. Precedence: bulk
  10.  
  11. >     How about a Pentium-specific Executor?  Any idea on how much 
  12. > speed-up?  With the 'Pentium' GCC, you get about 25% speed increase.  
  13.  
  14. I immediately thought, shouldn't matter since executor compiles the Mac
  15. code itself, and that's the code that takes most of the time.  But is that
  16. true?  How much time does executor spend in its various modules (screen
  17. update, instruction emulation, routine compilation, routine execution)?
  18.  
  19. Btw, the assembler output from gcc (and presumably djgcc) looks quite a
  20. bit more pentium-friendly if -fforce-mem is specified.  OTOH there seems
  21. to be a bug in -fforce-mem in the latest gcc snapshot, so it might not
  22. work.
  23.  
  24. --Arnt
  25.  
  26.