Exupery September 2006

exupery@lists.squeakfoundation.org

5 participants
6 discussions

Strongtalk and Exupery
by Bryce Kampjes 24 Sep '06

24 Sep '06

Hi David, It's great to see that the Strongtalk VM has been open sourced. Hopefully, it will be an asset to the community. Does the Strongtalk mailing list have a publicly available archive? One that doesn't require a yahoo sign-on? It would make it much easier for interested outsiders to follow what's going on. How does Strongtalk compare with current Java Hotspot VMs? They are also available with source for study (though not open sourced). I'm the primary author of Exupery, another attempt at fast execution technology for Smalltalk. Exupery is written in Smalltalk. The original design was to combine Self's dynamic inlining with a strong optimising compiler. For that the goal, I don't think we can afford to write in anything less productive than Smalltalk. That is still the goal but it's a long way off, Exupery is currently moving towards a 1.0 without full method inlining and without a strong optimiser. All the needed high risk features are there. Compile time is not the key issue for a dynamic compiler, pauses are. Compile time only becomes critical if you are stopping execution to compile. Exupery doesn't. Being normal Smalltalk like everything else pausing execution to compile is tricky. The trade offs to allow Exupery to be easily written in Smalltalk are the same as those required to allow long compile times for high grade optimisations. If you, or other Strongtalkers are interested in talking about compiler design please feel free to join Exupery's mailing list. Don't worry if you don't have time to study the source or play with it. Sharing experience would be valuable. Exupery is now about 4 years old, revisiting the design decisions with knowledgeable people would be useful, especially in an archived list. Exupery is another chance to keep the ideas and vision alive, if not the C++. The Exupery mailing list is here: http://lists.squeakfoundation.org/cgi-bin/mailman/listinfo/exupery Exupery's tiny benchmarks are: 1,176,335,439 bytecodes/sec; 16,838,438 sends/sec and with the interpreter: 228,367,528 bytecodes/sec; 7,241,760 sends/sec Which makes it currently much slower than Strongtalk for sends but the same speed for bytecodes. That's comparing against the numbers provided by Gilad via Dan's post to squeak-dev. Such a comparison is not fair as relative performance does vary greatly with architecture. Exupery is best on P4's, ok on Athlons, and least impressive on Pentium-Ms. The bytecode performance is the most interesting to me. Exupery does not yet do dynamic method inlining which explains Strongtalks strong send performance. Message inlining is not necessary for a 1.0. That the bytecode numbers are so close, and I know Exupery's weaknesses, is interesting. Exupery uses a colouring coalescing register allocator but also lives with Squeak's object memory and could do with a bit more tuning. I'm guessing Strongtalk's object memory is much cleaner and better designed for speed based on reading the Self papers. Did the Strongtalk team stop tuning for bytecode performance after they passed VisualWorks? Exupery has also recently been ported to Win 32 and Solaris 10 x86. Both ports were done by other people. Pre-built VMs will be available for both platforms in a few days. Bryce

2 4

[ANN] Squeak VM for Windows.
by Bryce Kampjes 23 Sep '06

23 Sep '06

Many thanks to Andrew Tween for porting Exupery to Windows. The VM is here: http://ftp.squeak.org/Exupery/vms/exupery-vm-0.09-win32.zip The installation instructions here: http://minnow.cc.gatech.edu/squeak/3945 Now include both Windows and Solaris. Does anyone feel like trying to compile it on the x86 Mac? The Mac should be easier than Windows unless Apple redefined the C calling conventions which I doubt. Bryce

1 0

trying to build VM for Win32
by Andrew Tween 22 Sep '06

22 Sep '06

Hi Bryce and all, I am trying to build an Exupery VM for Win32. Question 1. - has anyone successfully done this? Question 2. - I get this error when compiling - gnu-interp.c: In function `exuperySetupMessageSend': gnu-interp.c:4567: `foo' undeclared (first use in this function) Looking at the corresponding method in Interpreter, I see... [(receiverClass < youngStart and: [picPosition ~= 0]) ifTrue: [self cCode: '*(int *) (picPosition + 12) = foo->receiverClass' What is 'foo' ? Does this give any hints to what I am doing wrong? I am using 3.8-6665full.image, VMMaker-wbk.35, (I installed this first) Exupery-wbk.195 (it also fails with .197), I got the vm source from http://squeakvm.org/svn/squeak/tags/unix-3.7-7 and I installed the files from http://www.kampjes.demon.co.uk/vms/exuperyVMFiles.tz Cheers, Andy

3 11

[Ann] Exupery VM for Solaris10/x86
by Patrick Mauritz 18 Sep '06

18 Sep '06

Hello people, with some help from Bryce, I've built the VM with Exupery plugin on Solaris 10/x86 and created a package. You can get it from http://ftp.squeak.org/Exupery/vms/exupery-vm-0.09-solaris.pkg and install it with pkgadd. The VM is based on unix-3.7-7 and contains everything except AsynchFilePlugin (which won't work on Solaris), MiscPrimitivePlugin (which won't build), and all the plugins not available on unix platforms, of course. It's also built without iconv support, as squeak depends on gnu iconv behaviour ("MACINTOSH" table doesn't exist, "MAC" table, given as alternative, won't cooperate with utf-8, so is useless, too). Also note that the package is built on Solaris 10u2 (06/06), so it might, or might not work on older releases. The VM passed all but a few tests in Exupery's test suite, some of which are known-broken. You still have to get yourself an image and the smalltalk part of Exupery to try it. Have Fun, Patrick Mauritz

3 2

[ANN] Exupery 0.09 now released.
by Bryce Kampjes 09 Sep '06

09 Sep '06

Exupery 0.09 includes a bug fix for a bug found by Yakov. Exupery 0.08 would stop the image being saved if it was run by a non-Exupery VM. There is a simple work around. Exupery 0.09 also includes dynamically inlined primitives. Primitive inlining is done by reading information out of the polymorphic inline caches. More primitives will need to be implemented before this provides a real world speed improvement but at least now Exupery will attempt to compile and inline primitives. Releases should be more frequent as Exupery moves towards a 1.0. It would be helpful if a few people began to experiment with using Exupery. Feedback on what needs to be done to provide a 1.0 release would be useful. At the moment both bug fixes and more primitives are required. Bryce

1 0

Preparing for Exupery 0.09.
by Bryce Kampjes 08 Sep '06

08 Sep '06

Yakov found a bug in Exupery 0.08. It is is not possible to save the image after loading Exupery using a stock VM. This is because Exupery now needs to clean up the code cache before saving and that code calls an Exupery primitive. The work-around is to modify the method Exupery>>shutDown shutDown self initialiseExupery Either comment out the call to initialiseExupery or use #on:do: to catch all errors. Because of this bug, I'm now preparing to release the current development code. After running the stress test, I've caught a bug in my current implementation of #at:. I'm going to fix this the release. Exupery 0.09 has dynamically inlined primitives. It can only compile #at: and #at:put: for pointer objects and instance variable getter primitives. More primitives are coming, but this bug deserves a release and I'd rather release the current development work than back-port a fix. A good back port would only touch two methods and a hacked fix would touch one method. Bryce

1 0

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

Exupery September 2006