October 2008 - squirrelfish-dev - lists.macosforge.org

JavaScript performance-sensitive areas
by David Majda 29 Oct '08

29 Oct '08

Hello, I am writing a series of articles about current JavaScript engines and techniques they use to improve JavaScript performance. In an introductory article of the series I am trying to explain the readers why fast JavaScript execution is a problem and where the biggest performance hurdles lie. I have some (hopefully educated) guesses, but I am very interested in opinions of the JavaScript engine developers. What are the most problematic areas or language features when trying to write a fast JavaScript implementation? Thanks for your answers in advance. David Majda -- Everyone gets everything he wants. --Captain Willard in Apocalypse Now Personal :: david(a)majda.cz :: www.majda.cz Work :: david.majda(a)impaladesign.cz :: www.impaladesign.cz

2 1

Please disable Safari's "Developer" menu when performance testing JavaScript
by Geoffrey Garen 20 Oct '08

20 Oct '08

Hi. As of r37730, JavaScript is now faster (up to 22% faster) if Safari's "Developer" menu is disabled. If you are performance testing JavaScript on the command-line, nothing has changed. But if you are performance testing JavaScript inside Safari, please disable Safari's "Developer" menu first. Cheers, Geoff

1 0

problem building SF from WebKit-r37605
by Dhillon, Chetan 20 Oct '08

20 Oct '08

Hi, I do not know if it is the right mail list to get help on this topic but if not please direct to the appropriate list if anyone knows of. I downloaded WebKit-r37605 from WebKit Nightly Builds link and have also installed all required tools, cygwin, WebKitSupportLibrary etc. I have been trying to build it on Windows XP machine for last couple of days but to no avail. It is failing at qmake. Build script seems to provide -r option to it but qmake does not like it. Please see logs below. Any help will be greatly appreciated. Here is my qmake version. $ qmake -v Qmake version: 1.07a (Qt 3.3.4) Qmake is free software from Trolltech AS. Thanks Chetan cdhillon@CDHILLON2 ~/My Documents/work/WebRunTime/WebKit-r37605 $ ./WebKitTools/Scripts/build-webkit Subroutine main::tmpnam redefined at /cygdrive/c/Documents and Settings/cdhillon /My Documents/work/WebRunTime/WebKit-r37605/WebKitTools/Scripts/webkitdirs.pm line 33 Subroutine main::tmpfile redefined at /cygdrive/c/Documents and Settings/cdhillon/My Documents/work/WebRunTime/WebKit-r37605/WebKitTools/Scripts/webkitdirs.pm line 33 Checking mod-date of WebKitSupportLibrary.zip... Current WebKitSupportLibrary is up to date Calling 'qmake CONFIG+=qt-port -r OUTPUT_DIR=/cygdrive/c/Documents and Settings/cdhillon/My Documents/work/WebRunTime/WebKit-r37605/WebKitBuild/Release /cygdriv e/c/Documents and Settings/cdhillon/My Documents/work/WebRunTime/WebKit-r37605/WebKit.pro CONFIG+=release CONFIG-=debug' in /cygdrive/c/Documents and Settings/cdhillon/My Documents/work/WebRunTime/WebKit-r37605/WebKitBuild/Release ***Unknown option -r Usage: qmake [mode] [options] [files] QMake has two modes, one mode for generating project files based on some heuristics, and the other for generating makefiles. Normally you shouldn't need to specify a mode, as makefile generation is the default mode for qmake, but you may use this to test qmake on an existing project Mode: -project Put qmake into project file generation mode In this mode qmake interprets files as files to be built, defaults to *.c; *.ui; *.y; *.l; *.ts; *.h; *.hpp; *.hh; *.H; *.hxx; *.cpp; *.cc; *.cxx; *.C -makefile Put qmake into makefile generation mode (default) In this mode qmake interprets files as project files to be processed, if skipped qmake will try to find a project file in your current working directory Warnings Options: -Wnone Turn off all warnings -Wall Turn on all warnings -Wparser Turn on parser warnings -Wlogic Turn on logic warnings Options: * You can place any variable assignment in options and it will be * * processed as if it was in [files]. These assignments will be parsed * * before [files]. * -o file Write output to file -unix Run in unix mode -win32 Run in win32 mode -macx Run in Mac OS X mode -d Increase debug level -t templ Overrides TEMPLATE as templ -tp prefix Overrides TEMPLATE so that prefix is prefixed into the va lue -help This help -v Version information -after All variable assignments after this will be parsed after [files] -cache file Use file as cache [makefile mode only] -spec spec Use spec as QMAKESPEC [makefile mode only] -nocache Don't use a cache file [makefile mode only] -nodepend Don't generate dependencies [makefile mode only] -nomoc Don't generate moc targets [makefile mode only] -nopwd Don't look for files in pwd [ project mode only] -norecursive Don't do a recursive search [ project mode only] Failed to setup build environment using qmake!

1 0

Native code generation
by Guillaume Borios 02 Oct '08

02 Oct '08

Hi All, Maybe my question will be naive, but has anyone an idea/opinion about using llvm as a medium to get multiple native architecture support in SF? I mean produce optimized (from high level) llvm byte code and let the already existing llvm JIT compilation tools produce the optimal native code for the target platform. As far as I can see, pros are : - platform specific low level optimizations and code generation already done, - performance vs "hand generated" assembly? - support for many platforms - no more need of a complex (to design) ABI in SF to support architectural differences of target platforms - can focus on high level optimizations only, - maintenance? - fill with your own ideas... cons may be (many "?" because I actually only know concepts of llvm, and never really played with it yet!) : - security, maturity, etc? - support for new architectures harder (need to develop new llvm backends, or do like today, losing the advantages of llvm) - support for different OS? - licensing? embedding code from external project? - time to compile? - is it possible to bridge generated code with already running code, in both directions to keep using precompiled generic routine? - performance vs "hand generated" assembly? - maintenance? - fill with your own ideas again... As I started to work on the basis of a ppc jit compiler, I realize it won't be an easy task to write and maintain good compilers for multiple (and very different) hw architectures... -Guillaume In case it was a good question, credits to a friend for the idea :)

2 1

Some naming thoughts
by Maciej Stachowiak 02 Oct '08

02 Oct '08

StructureID: It's more than just an ID now. Besides the renaming, it should perhaps be merged with PropertyMap since there isn't as clearly a separate role. As for names: - Structure Pro: "these two objects have the same structure" sounds good Con: - StructureDescriptor - Shape Pro: "these two objects have the same shape"; short & sweet; Con: maybe don't want to rip off Mozilla's name? VM/CTI => jit/NativeCodeGeneratorX86 or jit/MachineCodeGeneratorX86 masm/X86Assembler => jit/AssemblerX86 I'm pretty sure this is good, but "Machine" or "Native"? I like "Native" somewhat better, now that we use "JS" and "Host" to distinguish between things originally in JS and originally in C++. Thought: will we end up with a platform-independent core for NativeCodeGenerator or will it have to be rewritten from scratch for every CPU architecture? We could probably make more of the code and logic reusable by having helper functions that emit the code for high- level operations, similar to things like emitJumpSlowCaseIfNotJSCell, or even larger chunks like compileBinaryArithOp. Regards, Maciej

3 4