Squeak-dev May 2009

squeak-dev@lists.squeakfoundation.org

102 participants
131 discussions

Re: [Vm-dev] stack vm questions
by Eliot Miranda 14 May '09

14 May '09

Hi Jecel, you ask some pointed questions on the Stack & JIT Cog VMs that need good answers and I suspect are generally interesting so I'll answer to squeak-dev also. On Thu, May 14, 2009 at 4:36 AM, Jecel Assumpcao Jr <jecel(a)merlintec.com>wrote: > [snip] > Eliot wrote: > > Yes. In the JIT an interpreted frame needs an extra field to hold > > the saved bytecode instruction pointer when an interpreted frame > > calls a machine code frame because the return address is the "return > > to interpreter trampoline" pc. There is no flag word in a machine > > code frame. So machine code frames save one word w.r.t. the > > stack vm and interpreted frames gain a word. But most frames > > are machine code ones so most of the time one is saving space. > > Ok, so the JIT VM will still have an interpreter. Self originally didn't > have one but most of the effort that David Ungar put into the project > when it was restarted was making it more interpreter friendly. The > bytecodes became very similar to the ones in Little Smalltalk, for > example. > > Will images be compatible between the JIT VM and the Stack VM? At least in the first version of the JIT yes. There are a couple of restrictions. - the code that rebuilds the Character table in SystemDictionary>>recreateSpecialObjectsArray has to be replaced by code that simply copies the table. This is because the JIT caches the Character table's oop in the generated machine-code for the at: primitive - the use (e.g. in Pharo) of the SmallInteger primitives 1 through 17 on LargePositiveInteger must be replaced by the relevant large integer primitives, which will be included in both the Stack and JIT Cog VMs. But I do expect to evolve the object representation quite soon, and this will break image backward-compatibility altogether. But I will want to produce a Stack VM that is compatible with the new format, not just a JIT VM because platform spread is very important to maintain and it'll be a while before more ISAs than x86 will be supported. Or do you > expect the latter to not be used anymore once the JIT is available? I expect that an interpreted Stack VM will survive for a long time and be the standard VM on ISAs such as SPARC, Itanium, MIPS, while the JIT will be available probably in the following order IA32/x86, ARM32, x86-64, PowerPC. I had originally understood that the Stack VM would be compatible with > older images (since you divorce all frames on save and remarry them on > reload) but I had missed the detail of the different bytecodes for > variable instance access in the case of Context objects. No, the Stack VM is not compatible with older images. The Context VM with closure bytecode extensions (e.g. John's current 4.0 beta VM) is backward-compatible and is the bridge over which one can bring forward pre-closure images to the Stack VM and JIT VM which are Closure-only VMs. One could make the Stack VM run older images but it would be difficult (and hence probably buggy) and I think unnecessary. One can always run older images on older VMs for as long as OSs support the older executables (and that other VM technology makes that easier these days). One can migrate an older image forward to the Stack VM using the bootstrap and the Context VM with Closure extensions. Packaging technologies such as Monticello make images much less important these days. One should expect to be able to export one's code in the form of a package and load it into a newer image, perhaps with a little massage. So keeping those dusty decks running on up-to-date VMs is IMO both wasted effort and a hindrance to obtaining maximum performance on newer VMs. Best Eliot

1 0

[BUG][FIX] Bug in WeakMessageSend
by Juan Vuletich 14 May '09

14 May '09

Hi Folks, I found a bug in WeakMessageSend that generates a DNU because of the message receiver being collected and disappearing. The culprits are #ensureReceiver and #ensureReceiverAndArguments . These methods ensure nothing. It might happen (although not often) that the receiver or some argument is collected after this method is called, but before it is actually used. The problem happens in Squeak 3.10, Pharo-10243, and Cuis. Please take a look. Also take a look at my proposed solution. I'm not sure, maybe some will object keeping an artificial reference to an object that could have been collected otherwise, eve if it is just for the time of a message send... Anyway, comments welcome. Cheers, Juan Vuletich 'From Squeak3.7 of ''4 September 2004'' [latest update: #5989] on 12 May 2009 at 11:02:25 am'! !WeakMessageSend methodsFor: 'evaluating' stamp: 'jmv 5/12/2009 10:59'! value ^ arguments isNil ifTrue: [ self withEnsuredReceiverDo: [ :r | r perform: selector]] ifFalse: [ self withEnsuredReceiverAndArgumentsDo: [ :r :a | r perform: selector withArguments: a]]! ! !WeakMessageSend methodsFor: 'evaluating' stamp: 'jmv 5/12/2009 11:00'! valueWithArguments: anArray | argsToUse | "Safe to use, because they are built before ensureing receiver and args..." argsToUse _ self collectArguments: anArray. ^self withEnsuredReceiverAndArgumentsDo: [ :r :a | r perform: selector withArguments: argsToUse ]! ! !WeakMessageSend methodsFor: 'evaluating' stamp: 'jmv 5/12/2009 11:01'! valueWithEnoughArguments: anArray "call the selector with enough arguments from arguments and anArray" | args | args _ Array new: selector numArgs. args replaceFrom: 1 to: ( arguments size min: args size) with: arguments startingAt: 1. args size > arguments size ifTrue: [ args replaceFrom: arguments size + 1 to: (arguments size + anArray size min: args size) with: anArray startingAt: 1. ]. "args is safe to use, because they are built before ensureing receiver and args..." ^self withEnsuredReceiverAndArgumentsDo: [ :r :a | r perform: selector withArguments: args ]! ! !WeakMessageSend methodsFor: 'private' stamp: 'jmv 5/12/2009 10:54'! withEnsuredReceiverAndArgumentsDo: aBlock "Grab real references to receiver and arguments. If they still exist, evaluate aBlock." "Return if my receiver has gone away" | r a | r _ self receiver. r ifNil: [ ^self ]. "Make sure that my arguments haven't gone away" "arguments can not become nil. The extra care in #ensureReceiverAndArguments is wrong (BTW, the whole idea is wrong, that's why we do these new methods...)" a _ Array withAll: arguments. a with: shouldBeNil do: [ :arg :flag | arg ifNil: [ flag ifFalse: [ ^self ]] ]. aBlock value: r value: a! ! !WeakMessageSend methodsFor: 'private' stamp: 'jmv 5/12/2009 10:53'! withEnsuredReceiverDo: aBlock "Grab a real reference to receive. If still there, evaluate aBlock." "Return if my receiver has gone away" | r | r _ self receiver. r ifNil: [ ^self ]. aBlock value: r! ! WeakMessageSend removeSelector: #ensureReceiver! WeakMessageSend removeSelector: #ensureReceiverAndArguments!

2 6

Smalltalk Television beta
by Chris Cunnington 14 May '09

14 May '09

Hi I've created a web page to introduce people to Smalltalk. It can be found here http://www.smalltalktelevision.com Chris

3 2

Re: [squeak-dev] Error installing Omnibase
by marcelo Cortez 14 May '09

14 May '09

Mariano ,people Sorry i'm forgot to mention , yes win32 platform. best mdc --- El jue 14-may-09, Mariano Martinez Peck <marianopeck(a)gmail.com> escribió: > De: Mariano Martinez Peck <marianopeck(a)gmail.com> > Asunto: Re: [squeak-dev] Error installing Omnibase > Para: "The general-purpose Squeak developers list" <squeak-dev(a)lists.squeakfoundation.org> > Fecha: jueves, 14 de mayo de 2009, 10:26 am > > > On Thu, May 14, 2009 at 12:13 PM, > Sebastian Sastre <ssastre(a)seaswork.com> > wrote: > > > > > > > Hi Mariano, you need to > install FFI > first > It was Marcelo not me ;) I just thought the same and so > my questions haha > > > > > > > > De: > squeak-dev-bounces(a)lists.squeakfoundation.org > > [mailto:squeak-dev-bounces@lists.squeakfoundation.org] > En nombre de > Mariano Martinez Peck > Enviado el: Thursday, May 14, 2009 > 10:10 > Para: The general-purpose Squeak developers > list > Asunto: Re: [squeak-dev] Error installing > Omnibase > > > > > > On Thu, May 14, 2009 at 1:41 AM, > marcelo Cortez <jmdc_marcelo(a)yahoo.com.ar> > > wrote: > > > folks > > i've > tried to install omnibase in many squeak images > versions without > sucess, > > the error is: > > closeHandle: aHandle > > "Close an open Win32 object handle, freeing > any resources held > by it. > Once closed a handle is no longer > valid. Answer whether the function > succeeds. > See Win32 SDK help for more > information. > > BOOL CloseHandle( > > HANDLE hObject > // handle of object to close > > );" > > <> expected ->apicall: > ulong ''CloseHandle'' (long) > module:''kernel32.dll''> > > ^self externalCallFailed' > > > > How are you installing it ? from Universes, SqueakMap or > Monticello? > > Do you have FFI installed ? > > Are you in Windows > ? > > Cheers, > > Mariano > > > > any > ideas? > tia. > > mdc > > btw, i use seaside image > > > > Yahoo! Cocina > Recetas prácticas y comida saludable > http://ar.mujer.yahoo.com/cocina/ > > > > > > > > > > > > -----Adjunto en línea a continuación----- > > > Yahoo! Cocina Recetas prácticas y comida saludable http://ar.mujer.yahoo.com/cocina/

2 1

Re: [squeak-dev] Error installing Omnibase
by marcelo Cortez 14 May '09

14 May '09

Mariano --- El jue 14-may-09, Mariano Martinez Peck <marianopeck(a)gmail.com> escribió: > De: Mariano Martinez Peck <marianopeck(a)gmail.com> > Asunto: Re: [squeak-dev] Error installing Omnibase > Para: "The general-purpose Squeak developers list" <squeak-dev(a)lists.squeakfoundation.org> > Fecha: jueves, 14 de mayo de 2009, 10:09 am > > > On Thu, May 14, 2009 at 1:41 AM, > marcelo Cortez <jmdc_marcelo(a)yahoo.com.ar> > wrote: > > > > folks > > > > i've tried to install omnibase in many squeak images > versions without sucess, > > > > the error is: > > > > closeHandle: aHandle > > "Close an open Win32 object handle, > freeing any resources held by it. > > Once closed a handle is no longer valid. Answer > whether the function > > succeeds. See Win32 SDK help for > more information. > > > > BOOL CloseHandle( > > HANDLE hObject // > handle of object to close > > );" > > > > <> expected ->apicall: ulong > ''CloseHandle'' (long) > module:''kernel32.dll''> > > ^self externalCallFailed' > > > > How are you installing it ? from Universes, SqueakMap or > Monticello? Yes all of them , the same error every time > > Do you have FFI installed ? Yes and try with ALien , whitout success. > > Are you in Windows ? > > Cheers, > > Mariano Best MDC > Yahoo! Cocina Recetas prácticas y comida saludable http://ar.mujer.yahoo.com/cocina/

1 0

Error installing Omnibase
by marcelo Cortez 14 May '09

14 May '09

folks i've tried to install omnibase in many squeak images versions without sucess, the error is: closeHandle: aHandle "Close an open Win32 object handle, freeing any resources held by it. Once closed a handle is no longer valid. Answer whether the function succeeds. See Win32 SDK help for more information. BOOL CloseHandle( HANDLE hObject // handle of object to close );" <> expected ->apicall: ulong ''CloseHandle'' (long) module:''kernel32.dll''> ^self externalCallFailed' any ideas? tia. mdc btw, i use seaside image Yahoo! Cocina Recetas prácticas y comida saludable http://ar.mujer.yahoo.com/cocina/

3 4

Questions regarding FFI
by Ricardo Moran 14 May '09

14 May '09

Sorry, I sent it with the wrong subject. Here I go again: On Wed, May 13, 2009 at 3:41 PM, Ricardo Moran <richi.moran(a)gmail.com>wrote: > Thanks for the reply, Mariano! I tried your advice but I ended up with some > "Could not coerce arguments" errors and crashing the VM several times :P. > However, I tried something like this: > > IplImage>>free > SqueakCV apiReleaseImage: handle asByteArrayPointer. > > SqueakCV>>apiReleaseImage: doublePointer > <apicall: void 'cvReleaseImage' (byte*) module: 'cxcore110.dll'> > ^ self externalCallFailed > > > It appears to be working. I don't know if this is the right way to do it, > though. I don't know either if overriding #free is ok and it would not bring > problems later... > > >> >* 2) To finalize objects the OpenCV library defines some functions such*>* as cvReleaseImage, cvReleaseMat, etc. As far as I can tell, these functions*>* receive a pointer-to-pointer as argument. How can I declare such a method in*>* Squeak? Having one object, how can I get the pointer to that object? Should*>* I use the #getHandle method?*>** >> In OpenDBX we have a functions like this for example: >> >> long odbx_init(odbx_t**, char*, char*, char*) >> >> This is how I declare it: >> >> apiInitialize: handle backend: backend host: host port: port >> <cdecl: long 'odbx_init' (ulong* char* char* char*) module: 'opendbx'> >> ^self externalCallFailed >> >> Remember a pinter to a pointer is a long :) >> >> And then in the code I do something like this: >> >> handleArray := WordArray with: 0. >> err := OpenDBX current >> apiInitialize: handleArray >> backend: self backend >> host: aConnection settings host >> port: aConnection settings port. >> >> Notice the WordArray. >> >> I have no idea of 1) and 3) sorry. >> >> I hope this helps. If this do help, perhaps we can put it in FFI wiki page. >> >> Cheers, >> >> Mariano >> >> >

2 1

[ANN][Dev-Images] May 2009 versions
by Damien Cassou 14 May '09

14 May '09

Hi, I've just uploaded new dev and web images based on Squeak3.9.1-7075, Squeak3.10.2-7179 and Pharo0.1-10303. More information and download: http://damiencassou.seasidehosting.st/Smalltalk/squeak-dev Pharo-based images can be downloaded at: http://pharo-project.org/download Download statistics at http://damiencassou.dabbledb.com/publish/dev-imagesdownloads Changes in this version: --------------------------- - Pharo images are based on version 10303 - OmniBrowser changes (by David Röthlisberger) - Heavily improved performance for browser updates after recompilations (Alex, David) - Reorganized menus for packages, classes, methods - New history button providing list of forward/backard history items (Simon) - Fixed protocol browser (Michal) - Minor layout improvings - Minor bug corrections Contained packages: ------------------------ This image contains the following packages (some packages are only in the squeak-web image): Aida version 5.6 Algernon version 1.2 AST version 164 AutomaticMethodCategorizer version 0.25 AutomaticMethodCategorizerOB version 0.2 DynamicBindings version 2.7 eCompletion version 0.102 eCompletion-Traits version 0.1 eCompletionOmniBrowser version 0.5 Installer version 4.3 Installer-Core version 4.3.313 Installer-Formats version 4.3.3 Installer-Scripts version 4.3.8 KomHttpServer version 7.0.45 KomServices version 1.19 Magritte-Model version 1.1.1.340 Magritte-Seaside version 1.1.1.284 Magritte-Tests version 1.1.1.149 OB-Enhancements version 0.327 OmniBrowser version 0.447 OmniBrowser-Algernon version 0.5 OmniBrowser-Full version 0.27 OmniBrowser-Morphic version 0.88 OmniBrowser-Refactory version 145 OmniBrowser-Regex version 0.18 OmniBrowser-Standard version 0.406 Pier version current Pier-Blog version 1.1.1.108 Pier-Documents version 1.1.1.8 Pier-EditorEnh version 1.1.1.17 Pier-Model version 1.1.1.271 Pier-Seaside version 1.1.1.340 Pier-Security version 1.1.1.121 Pier-Tests version 1.1.1.117 Polymorph EventEnhancements version 1.1 Polymorph Geometry version 1.1 Polymorph TaskbarIcons version 1.0 Polymorph ToolBuilder version 1.3 Polymorph Tools Diff version 1.3 Polymorph Widgets version 1.4 Refactoring Core version 48 Refactoring Spelling version 5 RoelTyper version 0.60 RSRSS2 version 2.8.3.11 Scriptaculous version 2.8.3.242 ScriptManager version 0.6 Seaside version 2.8.3.572 Seaside-Adapters-Core version pmm.2.mcz Seaside-Adapters-Swazoo version 9 Seaside-Squeak-Adapters version pmm.3.mcz SeasideAdaptersCompatibility version pmm.1.mcz Shout version 3.15-tween.72 ShoutWorkspace version 1-tween.4 SmaCC runtime version 13 Sport version 2.31 Squeak dev packages version 0.28 Squeak web packages version 0.2 SUnit-improved version 4.0.118 SUnitGUI-improved version 4.0.50 Swazoo version 2.1 Universes version 52 Universes OmniBrowser version 0.38 VBRegex version 1.9 YAXO version 14 -- Damien Cassou http://damiencassou.seasidehosting.st "Lambdas are relegated to relative obscurity until Java makes them popular by not having them." James Iry

1 0

Re: Questions regarding FFI
by kemal asad-syed 14 May '09

14 May '09

I don't have much of answers for you : but for > 3) Some OpenCV functions are declared as inline functions. I haven’t > been able to access those functions from Squeak. Is this an expected > behavior? for FFI to work you have to export your functions in the shared library, so they can be called in your shared library, and this is process is platform dependent. beleive you are compiling with gcc on gnu/Linux please confirm if not please let me know the platform . I don't see why inline function will be exported by default they are "just" used in the preprocess and at compilation time they are not regular functions. So what you found should be usual behavior. good luck. Kemal Depanage informatique http://je-resous.fr/

1 0

SqueakDocs an iphone electronic book hits App store.
by John M McIntosh 14 May '09

14 May '09

I was pleased to get notification that my Squeak & Pharo Smalltalk electronic books went up for sale on the App Store this afternoon. Sadly WikiServer http://www.mobilewikiserver.com a much more complex application is still stuck in the review process. Basically the two books are Seaside apps that use an altered Seaside class browser to give you a READ-ONLY view of all the source code from a 3.10.x Squeak web image, or Pharo web dev image of late April 2009. Squeak -> http://itunes.apple.com/WebObjects/MZStore.woa/wa/viewSoftware?id=315289375… Pharo -> http://itunes.apple.com/WebObjects/MZStore.woa/wa/viewSoftware?id=315254856… Unfortunately because I have to recoup the $$$ I send Apple each year for developer fees and deal with support issues means they are not free. I will discuss these apps and how they were developed in detail at ESUG 09. http://www.esug.org/Conferences/2009/Current+List+of+Talks People who need a promo coupon to review & use the app without having to pay the iTunes store can email me privately, I have a limited number of these to hand out. -- = = = ======================================================================== John M. McIntosh <johnmci(a)smalltalkconsulting.com> Twitter: squeaker68882 Corporate Smalltalk Consulting Ltd. http://www.smalltalkconsulting.com = = = ========================================================================

1 0

← Newer
1
...
5
6
7
8
9
10
11
...
14
Older →

Jump to page: