swarm-support
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: MessageProbe speed-ups


From: Pietro Terna
Subject: Re: MessageProbe speed-ups
Date: Sat, 14 Jul 2001 21:21:56 +0200

        Wonderful! Pietro

At 12.17 14/07/2001 -0600, you wrote:
>>>>> "PT" == Pietro Terna <address@hidden> writes:

PT> in "latest" the dll files are always dated July 7

The DLLs are up to date with today's snapshot.

  ftp://ftp.swarm.org/pub/swarm/src/testing/swarm-2001-07-14.tar.gz
  ftp://ftp.swarm.org/pub/swarm/binaries/w32/latest/javaswarm.dll.gz
                                                    libkaffeswarm.dll.gz
                                                    swarm.dll.gz
                                                    swarm.jar.gz

The speed-up you see ought to be more dramatic than I indicated in
that e-mail because today's changes remove most Java -> Objective C
object lookup costs (throughout Swarm).  With Kaffe on
sparc-sun-solaris2.7 the MessageProbe change together with the lookup
elimination results in batch runs for Java heatbugs that are more than
twice as fast.  On i686-pc-linux-gnu system, the speed-up is larger.
[Interestingly, the lookup elimination change _really_ wins with Kaffe
on Solaris but not so much with Sun JDK on Solaris.  However, on Intel
there isn't so much difference.]

More performance work is underway..

                  ==================================
   Swarm-Support is for discussion of the technical details of the day
   to day usage of Swarm.  For list administration needs (esp.
   [un]subscribing), please send a message to <address@hidden>
   with "help" in the body of the message.


                 ==================================
  Swarm-Support is for discussion of the technical details of the day
  to day usage of Swarm.  For list administration needs (esp.
  [un]subscribing), please send a message to <address@hidden>
  with "help" in the body of the message.



reply via email to

[Prev in Thread] Current Thread [Next in Thread]