Vm-dev August 2011

vm-dev@lists.squeakfoundation.org

40 participants
111 discussions

Some problems with latest VM update
by Igor Stasenko 27 Aug '11

27 Aug '11

1. Huge allocation crashing Cog VM on Unix 0xbfccc72c M [] in AllocationTest>testOutOfMemorySignal 2030998324: a(n) AllocationTest You can see the full log at: https://ci.lille.inria.fr/pharo/view/Cog/job/Cog-Unix%20Tests/22/console 2. Stack VMs has a compilation errors: Undefined symbols: "_noInlineTemporaryin", referenced from: _stObjectat in gcc3x-interp.c.o You can see the full log at: https://ci.lille.inria.fr/pharo/view/Cog/job/Cog-Unix%20Tests/22/console -- Best regards, Igor Stasenko AKA sig.

2 1

VMs don't compile on Lion
by Sean P. DeNigris 27 Aug '11

27 Aug '11

To save Lion users the trouble... After banging my head against the wall all week at ESUG (thank you Esteban, Igor, and Elliot for all the support), it seems that the Cog family of VMs (tried every variant of Jit/Stack & Carbon/Cocoa) do not compile on OS X Lion. I performed the same steps as both Igor (who succeeded in Snow Leopard) and Jenkins, with errors in during make. -- View this message in context: http://forum.world.st/VMs-don-t-compile-on-Lion-tp3771938p3771938.html Sent from the Squeak VM mailing list archive at Nabble.com.

3 3

Build error on Stack VM
by Sean P. DeNigris 26 Aug '11

26 Aug '11

After these steps: 1. Fresh Pharo 1.3 image 2. MetacelloRepository ConfigurationOfCog 3. ConfigurationOfCog project latestVersion load. 4. CocoaIOSCogStackConfig generateWithSources 5. cd build 6. cmake . 7. make I got the following error: > CMake Error at CroquetPlugin/CMakeLists.txt:17 (add_library): > Cannot find source file: > > > /Users/sean/Squeak/vm/seandenigris-cog-blessed/platforms/Cross/plugins/CroquetPlugin/tribox.c > > Tried extensions .c .C .c++ .cc .cpp .cxx .m .M .mm .h .hh .h++ .hm .hpp > .hxx .in .txx > I commented out the offending line in CMakeLists.txt and then got the following: I got this error: > Linking CXX executable > /Users/sean/Squeak/vm/seandenigris-cog-blessed/results/StackVM.app/Contents/MacOS/StackVM > Undefined symbols for architecture i386: > "_getStackPointer", referenced from: > _sqGetInterpreterProxy in sqVirtualMachine.c.o > "_signalNoResume", referenced from: > _sqGetInterpreterProxy in sqVirtualMachine.c.o > ld: symbol(s) not found for architecture i386 > collect2: ld returned 1 exit status > make[2]: *** > [/Users/sean/Squeak/vm/seandenigris-cog-blessed/results/StackVM.app/Contents/MacOS/StackVM] > Error 1 > make[1]: *** [CMakeFiles/StackVM.dir/all] Error 2 > make: *** [all] Error 2 > What do I do now? Thanks. Sean -- View this message in context: http://forum.world.st/Build-error-on-Stack-VM-tp3770884p3770884.html Sent from the Squeak VM mailing list archive at Nabble.com.

2 2

The serverMode flag and some thoughts on event polling
by Dimitry Golubovsky 26 Aug '11

26 Aug '11

Hi, [this is not really about VM alone, but rather how image and VM interact - hence posting to multiple lists] In the process of porting Cog to Android, I observed the following pattern. If a scrollbar's "up" or "down" button is held pressed (prolonged tap on a button) for few seconds, the VM freezes, and gets killed by Android. Digging down the code got me to the WorldState>>interCyclePause: method. When there are morphs requesting stepping by shorter than certain time interval, the process responsibke for running step methods keeps looping without giving up to lower priority processes (and the host OS). ScrollBar seems to decrease the stepping interval over time to simulate accelerated scrolling. In order to port Cog to Android, I turned it into an event-driven VM based on ideas found in [1]. The same was done by Andreas for his earlier Classic VM port. My approach is a bit different such as the longjmp call is made as part of the relinquishProcessor primitive. Such primitive is called only from the lowest priority (idle) process which only can run when there are no external events to deal with. The event-driven VM needs to exit from the interpreter to the host process in order to obtain any input events. Thus, non-interrupted polling just deadlocks the interpreter as it cannot exit to the host process to get more events, yet keeps polling for them. The serverMode flag [2] (aka Mantis #6581) came to rescue. With such flag set to true, polling for events always yields CPU for 50ms (which in the Android case is meaningless anyway: I would not give the VM idle timer more often than 10 times per sec, so 50ms granularity is just not handleable). Anyway this does the trick as the lowest priority process gets its CPU time, and exits the interpreter to the host process. With serverMode = true I can tap a scroller button as long as I want, and even the text cursor in a workspace keeps blinking which means that stepping works and VM is alive. BTW in Android, the interpreter will be called as soon as any user action is intercepted, so responsiveness is not degraded regardless of the delay introduced by serverMode. I have found at least one place (this is in PharoCore) where serverMode is not honored: PolygonMorph class >> fromHand: (this method creates a polygon drawn by the mouse) This method has code pieces (repeating multiple times btw - why is it not a separate method?) [Sensor anyButtonPressed] whileFalse: [self currentWorld displayWorldSafely; runStepMethods]. which is clearly grabbing the event sensor to the polygon morph while keeping other morphs stepped, so the VM does not seem dead entirely. In my case though executing PolygonMorph fromHand: defaultHand freezes the VM immediately as this is uninterrupted polling again. While I'll try to experiment with inserting interCyclePause somewhere in these loops, there are broader questions on how event polling is done these days: 1. Are there other morphs in packages not included in PharoCore where event grabbing is done such way? 1a. Is this possible in general to have such functionality properly embedded in WorldState instead of having it in a morph itself, if event grabbing is needed? 2. Are there images in active development (other than PharoCore, and Squeak) where event polling is done without an option to relinquish processor? In fact, this VM freezing on stepping morphs was a big issue, and it mainly kept me from moving towards the public release. Now that it seems to be solved (unless there is a flaw in my logic above) it takes me closer to finishing the Cog for Android port. Thanks. ------------------------------------------ [1] http://lists.squeakfoundation.org/pipermail/vm-dev/2009-November/003385.html [2] http://bugs.squeak.org/view.php?id=6581 -- Dimitry Golubovsky Anywhere on the Web

3 2

Missing method in VMMaker-oscog-IgorStasenko.122
by Dimitry Golubovsky 26 Aug '11

26 Aug '11

Igor Stasenko wrote: > So, the fix is quite simple: > move method (noInlineTemporary: offset in: theFP) > from CoInterpreter class higher to StackInterpreter Yes, I fixed it locally on my computer like this, it works. Thanks. -- Dimitry Golubovsky Anywhere on the Web

1 0

Missing method in VMMaker-oscog-IgorStasenko.122
by Dimitry Golubovsky 26 Aug '11

26 Aug '11

Hi, Stack interpreter does not compile because the method noInlineTemporary: offset in: theFP defined only in CoIntepreter, yet it is called from StackInterpreter: ------------------------------------------------------------------------------------------ stObject: array at: index "Return what ST would return for <obj> at: index." | hdr fmt totalLength fixedFields stSize | <inline: false> hdr := objectMemory baseHeader: array. fmt := objectMemory formatOfHeader: hdr. totalLength := objectMemory lengthOf: array baseHeader: hdr format: fmt. fixedFields := objectMemory fixedFieldsOf: array format: fmt length: totalLength. (fmt = 3 and: [self isContextHeader: hdr]) ifTrue: [stSize := self fetchStackPointerOf: array. ((self oop: index isGreaterThanOrEqualTo: 1) and: [(self oop: index isLessThanOrEqualTo: stSize) and: [self isStillMarriedContext: array]]) ifTrue: [^self noInlineTemporary: index - 1 in: (self frameOfMarriedContext: array)]] " <---- here" ifFalse: [stSize := totalLength - fixedFields]. ((self oop: index isGreaterThanOrEqualTo: 1) and: [self oop: index isLessThanOrEqualTo: stSize]) ifTrue: [^self subscript: array with: (index + fixedFields) format: fmt] ifFalse: [self primitiveFailFor: PrimErrBadIndex. ^ 0] ------------------------------------------------------------------------------------------ Thanks. -- Dimitry Golubovsky Anywhere on the Web

2 2

Re: [Pharo-project] [Vm-dev] Cog in Jenkis has brokken finalization
by Mariano Martinez Peck 26 Aug '11

26 Aug '11

On Thu, Aug 25, 2011 at 5:40 PM, Marcus Denker <marcus.denker(a)inria.fr>wrote: > > On Aug 25, 2011, at 5:34 PM, Igor Stasenko wrote: > > > On 25 August 2011 14:38, Mariano Martinez Peck <marianopeck(a)gmail.com> > wrote: > >> > >> Hi guys. The linux Cog VM from hudson has broken the finalization. Just > run the WeakRegistryTest and you will see the broken test. It might > happened because of the ephemerons because Eliot's VM works correctly. > >> Even more, the VM from Jenkis is correct. So....I would remove the > broken one from Hudson. > >> > > > > I migrated unix and windows VM jobs to Jenkins. > > Please use VMs built on Jenkins servers. > > > > Can we delete the Hudson projects? Just to avoid confusion? > > That was all my email about in fact > -- > Marcus Denker -- http://marcusdenker.de > > > -- Mariano http://marianopeck.wordpress.com

2 1

VM Maker: VMMaker-dtl.251.mcz
by squeak-dev-noreply＠lists.squeakfoundation.org 26 Aug '11

26 Aug '11

Dave Lewis uploaded a new version of VMMaker to project VM Maker: http://www.squeaksource.com/VMMaker/VMMaker-dtl.251.mcz ==================== Summary ==================== Name: VMMaker-dtl.251 Author: dtl Time: 25 August 2011, 10:57:54 am UUID: 68ce54b1-a53b-40b2-aea9-870f62357f87 Ancestors: VMMaker-dtl.250 VMMaker 4.7.8 Abstract methods should not be translated, so treat methods with #subclassResponsibility just like #doNotTranslate or <doNotTranslate>. Add SlangTest tests to verify. Fix cut and paste errors in SlangTest from prior update. Note to self: Consider refactoring NewObjectMemory/ObjectMemory to eliminate overrides, marking abstract methods with #subclassResponsibility.

1 0

VM Maker: VMMaker-dtl.250.mcz
by squeak-dev-noreply＠lists.squeakfoundation.org 26 Aug '11

26 Aug '11

Dave Lewis uploaded a new version of VMMaker to project VM Maker: http://www.squeaksource.com/VMMaker/VMMaker-dtl.250.mcz ==================== Summary ==================== Name: VMMaker-dtl.250 Author: dtl Time: 25 August 2011, 9:33:50 am UUID: 5b1ed008-33c0-4af9-a4db-e77a321ed547 Ancestors: VMMaker-dtl.249 VMMaker 4.7.7 Adopt <doNotGenerate> support from oscog. Implement Object>>doNotGenerate equivalent of <doNotGenerate> Add SlangTest unit tests and support class methods to verify the above. @eliot - oddly harmless co-dependent bugs in CCodeGenerator>> addMethodFor:selector: which relies on the unspecified return value of #addMethod (thus a CCodeGenerator, not a TMethod), then attempts to disable inlining of the method by sending #inline: to the "TMethod" (actually a CCodeGenerator at this point), but #inline: is actually a no-op in Object, so nothing happens at all. Editor's note: Avoid use of #doNotGenerate for methods that shadow support code macros, e.g. marking pointerForOop: as a doNotGenerate method may conflict with an actual implementation of pointerForOop: that is meant for translation in class MemoryAccess. Ancestors: VMMaker-dtl.249 Added, Modified, Deleted vs. VMMaker-dtl.249: M CCodeGenerator M CCodeGenerator>>addClass: A CCodeGenerator>>addMacro:for: A CCodeGenerator>>addMethodFor:selector: A CCodeGenerator>>checkDeleteVariable: A CCodeGenerator>>compileToTMethodSelector:in: M CCodeGenerator>>initialize A CCodeGenerator>>removeMethodForSelector: A CCodeGenerator>>removeVariable: A Object class>>prepareToBeAddedToCodeGenerator: A Object>>doNotGenerate A SlangTest>>testInterpDoNotdoNotGenerateByMethod A SlangTest>>testInterpDoNotdoNotGenerateByPragma A SlangTest>>testPluginDoNotdoNotGenerateByMethod A SlangTest>>testPluginDoNotdoNotGenerateByPragma A SlangTest>>testSSIPDoNotdoNotGenerateByMethod A SlangTest>>testSSIPDoNotdoNotGenerateByPragma A SlangTestSupportInterpreter class>>requiredMethodNames A SlangTestSupportInterpreter>>doNotGenerateByMethod A SlangTestSupportInterpreter>>doNotGenerateByPragma A SlangTestSupportPlugin class>>requiredMethodNames A SlangTestSupportPlugin>>doNotGenerateByMethod A SlangTestSupportPlugin>>doNotGenerateByPragma A SlangTestSupportSSIP class>>requiredMethodNames A SlangTestSupportSSIP>>doNotGenerateByMethod A SlangTestSupportSSIP>>doNotGenerateByPragma A TMethod>>hasDoNotGenerateStatement M VMMaker class>>versionString

1 0

Cog in Jenkis has brokken finalization
by Mariano Martinez Peck 25 Aug '11

25 Aug '11

Hi guys. The linux Cog VM from hudson has broken the finalization. Just run the WeakRegistryTest and you will see the broken test. It might happened because of the ephemerons because Eliot's VM works correctly. Even more, the VM from Jenkis is correct. So....I would remove the broken one from Hudson. Cheers -- Mariano http://marianopeck.wordpress.com

2 1

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

Vm-dev August 2011