home *** CD-ROM | disk | FTP | other *** search
/ Executor 2.0 / executorv2.0.iso / pc / dos / extra / docs / maillist / text / archive.95 / text0638.txt < prev    next >
Encoding:
Internet Message Format  |  1996-03-31  |  996 b 
  1. To: executor@nacm.com
  2. Subject: Re: 486 optmization
  3. In-Reply-To: <Pine.LNX.3.91.950427073305.2351A-100000@zeus.achilles.net>
  4. Message-Id: <Pine.LNX.3.91.950427200945.14728B-100000@popcorn.troll.no>
  5. Mime-Version: 1.0
  6. Content-Type: TEXT/PLAIN; charset=US-ASCII
  7. Sender: owner-executor@nacm.com
  8. Precedence: bulk
  9.  
  10. >     How about a Pentium-specific Executor?  Any idea on how much 
  11. > speed-up?  With the 'Pentium' GCC, you get about 25% speed increase.  
  12.  
  13. I immediately thought, shouldn't matter since executor compiles the Mac
  14. code itself, and that's the code that takes most of the time.  But is that
  15. true?  How much time does executor spend in its various modules (screen
  16. update, instruction emulation, routine compilation, routine execution)?
  17.  
  18. Btw, the assembler output from gcc (and presumably djgcc) looks quite a
  19. bit more pentium-friendly if -fforce-mem is specified.  OTOH there seems
  20. to be a bug in -fforce-mem in the latest gcc snapshot, so it might not
  21. work.
  22.  
  23. --Arnt
  24.  
  25.