Vm-dev December 2010

vm-dev@lists.squeakfoundation.org

42 participants
109 discussions

VM Maker: VMMaker-dtl.209.mcz
by squeak-dev-noreply＠lists.squeakfoundation.org 12 Dec '10

12 Dec '10

Dave Lewis uploaded a new version of VMMaker to project VM Maker: http://www.squeaksource.com/VMMaker/VMMaker-dtl.209.mcz ==================== Summary ==================== Name: VMMaker-dtl.209 Author: dtl Time: 11 December 2010, 5:45:49 am UUID: 8a05bb12-7b76-449f-8042-b0d9e2315bea Ancestors: VMMaker-dtl.208 VMMaker 4.4.0 Update interpreter primitives to match oscog. Add #primitiveAdoptInstance from oscog, but change it to call #flushMethodCache rather than #flushAtCache (which is specific to the StackInterpreter). Note - suggest implementing #flushMethodCache in StackInterpreter for compatibility. Add large integer (64-bit) numbered primitives from oscog. Remove #primitiveDoNamedPrimitiveWithArgs. Remove #primitiveObsoleteIndexedPrimitive. Remove the obsoleteNamedIndexPrimitiveTable and obsoleteIndexedPrimitiveTable from Interpreter. Adopt #initializePrimitiveTable from oscog, except for VM implementor primitives 250 - 253 which are different in the platform support code.

1 0

Cog console messages starting Squeak4.2
by Ken G. Brown 12 Dec '10

12 Dec '10

I'm noticing these console messages when starting up a fresh Squeak4.2-10548-alpha.image on Cog.app 4.0.0 from http://www.mirandabanda.org/files/Cog/VM/VM.r2316/ on MacOS X 10.6.5 10/12/11 2:08:32 PM [0x0-0x283283].com.teleplace.Teleplace[6349] mprotect(x,y,PROT_READ | PROT_WRITE): Invalid argument >From Activity Monitor Croquet Threads 3 Real Memory 27.9 MB Intel Virtual Memory 50.1 MB Private memory 2.4 MB shared memory 28.4 MB ====== System Total Memory 8 GB Free memory 4.41 GB Wired 657.2 MB Active 1.27 GB Inactive 1.68 GB Used 3.59 GB VM size 148.61 GB page ins 5.58 GB Page outs 668 GB Swap used 0 MB Are the messages anythig to worry about? Ken G. Brown

3 2

Re: Reproducible way to crash Cog ( [Lsehub-staff] Strange behavior )
by Igor Stasenko 09 Dec '10

09 Dec '10

On 8 December 2010 22:41, Benjamin <benjamin.vanryseghem.pharo(a)gmail.com> wrote: > > On Dec 8, 2010, at 10:28 PM, Igor Stasenko wrote: > >> Hello, Eliot >> >> today i tried to help and see what is strange (see original message >> below).. and followed the instructions, but instead of 'strange' i got >> VM crash. ;) >> >> I tried multiple Mac Cog VMs (one built by me, another, built by you), >> and both crashing during GC. >> >> To reproduce, take the latest pharo-core image >> (http://www.pharo-project.org/pharo-download/unstable-core) >> Then evaluate following: >> >> Gofer new >> squeaksource: 'Seed'; >> package: 'Seed-HaselBuilder'; >> load. >> >> >> Then evaluate >> HaselKernelBuilder new script. >> >> if next to that you do (if it not crashed already) >> >> Smalltalk garbageCollect. >> >> >> and VM will crash guaranteed (or become unresponsive). >> >> I didn't investigated deeply, but i found that culprit is most probably in >> HaselKernelBuilder>>copyClass: >> >> >> >> @ Benjamin. You have a mistake in your code (and that i think why it >> behaves 'strange') >> you setting the class of class to be the instance of 'metacopy' >> >> so, you have >> >> class copy -> metaclass copy basicNew -> metaclass copy >> >> ('->' here means 'instance of' ) >> >> while it should be >> >> class copy -> metaclass copy >> >> So >> >> cpy primitiveChangeClassTo: metaCopy basicNew. >> >> actually should be >> >> cpy primitiveChangeClassTo: metaCopy. >> > > In fact, this primitive take the class of the argument to change the class > of the receiver (yes I know, it's strange ...) ah, sorry, my fault. Then its should be ok to do like that. But i got another crash before getting to that line , so i didn't have change to test it. It seems something wrong happening during method copying. I never used this prim before, so i didn't knew that it takes the class of argument, instead an argument itself as a class to which replace receiver's class. > > The 'strange' behavior is also a crash from the VM, but not in the kernel > creation. > > > Thank you for your advice (and the beach volley ^^) > > Ben > > >> >> ---------- Forwarded message ---------- >> From: Benjamin <benjamin.vanryseghem(a)gmail.com> >> Date: 8 December 2010 17:15 >> Subject: [Lsehub-staff] Strange behavior >> To: RMoD private list <lsehub-staff(a)lists.gforge.inria.fr> >> >> >> Hello guys, >> >> I've a strange error when I'm trying to trace some objects, and even >> after spending the whole day on it, I still do not understand why, so >> if you have any idea, you're welcome :) >> >> >> Source : >> >> Gofer new >> squeaksource: 'Seed'; >> package: 'Seed-HaselBuilder'; >> load. >> >> >> Method : >> >> Then open a Transcript and evaluate >> >> HaselKernelBuilder new script. >> tracedObjects := Dictionary new. objects := Dictionary new. object := >> Object. HaselImageBuilder new trace: object in: objects from: nil >> tracedObjects: tracedObjects. >> >> >> Thank you in advance, >> >> >> Ben >> >> >> >> -- >> Best regards, >> Igor Stasenko AKA sig. > > -- Best regards, Igor Stasenko AKA sig.

4 5

Suspicious typing in Cog VM
by Aleksej Saushev 09 Dec '10

09 Dec '10

Hello! Types are inconsistent along the source, does Cog VM work on LP64 platforms? I need this to make it build on NetBSD/i386: --- platforms/unix/plugins/SoundPlugin/sqUnixSound.c.orig 2010-12-07 17:40:51.000000000 +0000 +++ platforms/unix/plugins/SoundPlugin/sqUnixSound.c @@ -99,12 +99,12 @@ int snd_AvailableSpace(void) return snd->snd_AvailableSpace(); } -int snd_InsertSamplesFromLeadTime(int frameCount, int srcBufPtr, int samplesOfLeadTime) +sqInt snd_InsertSamplesFromLeadTime(sqInt frameCount, void *srcBufPtr, sqInt samplesOfLeadTime) { return snd->snd_InsertSamplesFromLeadTime(frameCount, srcBufPtr, samplesOfLeadTime); } -int snd_PlaySamplesFromAtLength(int frameCount, int arrayIndex, int startIndex) +sqInt snd_PlaySamplesFromAtLength(sqInt frameCount, void *arrayIndex, sqInt startIndex) { return snd->snd_PlaySamplesFromAtLength(frameCount, arrayIndex, startIndex); } @@ -141,7 +141,7 @@ double snd_GetRecordingSampleRate(void) return snd->snd_GetRecordingSampleRate(); } -int snd_RecordSamplesIntoAtLength(int buf, int startSliceIndex, int bufferSizeInBytes) +sqInt snd_RecordSamplesIntoAtLength(void *buf, sqInt startSliceIndex, sqInt bufferSizeInBytes) { return snd->snd_RecordSamplesIntoAtLength(buf, startSliceIndex, bufferSizeInBytes); } @@ -150,7 +150,7 @@ int snd_RecordSamplesIntoAtLength(int bu void snd_Volume(double *left, double *right) { snd->snd_Volume(left, right); } void snd_SetVolume(double left, double right) { snd->snd_SetVolume(left, right); } -int snd_SetRecordLevel(int level) { return snd->snd_SetRecordLevel(level); } +void snd_SetRecordLevel(sqInt level) { return snd->snd_SetRecordLevel(level); } int snd_GetSwitch(int id, int captureFlag, int channel) { return snd->snd_GetSwitch(id, captureFlag, channel); } int snd_SetSwitch(int id, int captureFlag, int parameter) { return snd->snd_SetSwitch(id, captureFlag, parameter); } int snd_SetDevice(int id, char *name) { return snd->snd_SetDevice(id, name); } --- src/vm/gcc3x-cointerp.c.orig 2010-12-07 18:00:09.000000000 +0000 +++ src/vm/gcc3x-cointerp.c @@ -975,7 +975,7 @@ sqInt quickPrimitiveInstVarIndexFor(sqIn sqInt rawHeaderOf(sqInt methodPointer); void rawHeaderOfput(sqInt methodOop, void *cogMethod); sqInt readableFormat(sqInt imageVersion); -sqInt readImageFromFileHeapSizeStartingAt(sqImageFile f, sqInt desiredHeapSize, squeakFileOffsetType imageOffset); +sqInt readImageFromFileHeapSizeStartingAt(sqImageFile f, usqInt desiredHeapSize, squeakFileOffsetType imageOffset); sqInt remap(sqInt oop); static sqInt removeFirstLinkOfList(sqInt aList); EXPORT(sqInt) removeGCRoot(sqInt *varLoc); @@ -37959,7 +37959,7 @@ readableFormat(sqInt imageVersion) { */ sqInt -readImageFromFileHeapSizeStartingAt(sqImageFile f, sqInt desiredHeapSize, squeakFileOffsetType imageOffset) { +readImageFromFileHeapSizeStartingAt(sqImageFile f, usqInt desiredHeapSize, squeakFileOffsetType imageOffset) { DECL_MAYBE_SQ_GLOBAL_STRUCT sqInt anObject; sqInt aValue; -- HE CE3OH...

6 13

Compact class question
by Gabriel Hernán Barbuto 09 Dec '10

09 Dec '10

Hi I am working on writing an image to disk. If I understand correctly, the size in words field in the one word header must take into account the word for the header. I have a CompiledMethod instance that has 63 words of content plus at least one header word for a total of 64 words. I think that I should write a three word header for this instance even though its class is compact. Is this correct? Is it fine to have an instance of compact class that doesn't have a one word header? Thanks in advance. Gabriel

6 11

Reproducible way to crash Cog ( [Lsehub-staff] Strange behavior )
by Igor Stasenko 08 Dec '10

08 Dec '10

Hello, Eliot today i tried to help and see what is strange (see original message below).. and followed the instructions, but instead of 'strange' i got VM crash. ;) I tried multiple Mac Cog VMs (one built by me, another, built by you), and both crashing during GC. To reproduce, take the latest pharo-core image (http://www.pharo-project.org/pharo-download/unstable-core) Then evaluate following: Gofer new squeaksource: 'Seed'; package: 'Seed-HaselBuilder'; load. Then evaluate HaselKernelBuilder new script. if next to that you do (if it not crashed already) Smalltalk garbageCollect. and VM will crash guaranteed (or become unresponsive). I didn't investigated deeply, but i found that culprit is most probably in HaselKernelBuilder>>copyClass: @ Benjamin. You have a mistake in your code (and that i think why it behaves 'strange') you setting the class of class to be the instance of 'metacopy' so, you have class copy -> metaclass copy basicNew -> metaclass copy ('->' here means 'instance of' ) while it should be class copy -> metaclass copy So cpy primitiveChangeClassTo: metaCopy basicNew. actually should be cpy primitiveChangeClassTo: metaCopy. ---------- Forwarded message ---------- From: Benjamin <benjamin.vanryseghem(a)gmail.com> Date: 8 December 2010 17:15 Subject: [Lsehub-staff] Strange behavior To: RMoD private list <lsehub-staff(a)lists.gforge.inria.fr> Hello guys, I've a strange error when I'm trying to trace some objects, and even after spending the whole day on it, I still do not understand why, so if you have any idea, you're welcome :) Source : Gofer new squeaksource: 'Seed'; package: 'Seed-HaselBuilder'; load. Method : Then open a Transcript and evaluate HaselKernelBuilder new script. tracedObjects := Dictionary new. objects := Dictionary new. object := Object. HaselImageBuilder new trace: object in: objects from: nil tracedObjects: tracedObjects. Thank you in advance, Ben -- Best regards, Igor Stasenko AKA sig.

2 2

Error: sweep failed to find exact end of memory
by Mariano Martinez Peck 08 Dec '10

08 Dec '10

Hi, I am having this error quite frequently while working with the proxies for CompiledMethods. I don't understand....I am reaching a maximum in the memory? I checked both, the amount of Objects I create and the amount of memory and doesn't look that large. I am in mac and I am even doing -memory 1500m and also setting HeapSize in plist.info. Any hints? I serached in Google and I found a bug with signed and unsiged stuff...but i didn't understand and I am not sure which is the status today (the threads were quite old). Here is an example: sweep failed to find exact end of memory 594035904 Object>myDependents: 594035812 Object>breakDependents 594035720 Object>actAsExecutor 594035628 StandardFileStream>actAsExecutor 594035496 Object>executor 594035332 WeakRegistry>add: 594035240 >register: 594035148 StandardFileStream>register 594034104 StandardFileStream>open:forWrite: 594032588 >fileNamed: 594032496 FuelMareaSerializer>mareaUnserializeMethodFrom: 594032216 >mareaLoadCompiledMethodFor: 594032124 >mareaUninstallForProxy:ifNotInstalled: 594032032 >mareaUninstallForProxy: 594027820 >mareaHandleRun:with:in:forProxy: 594027640 AbstractCompiledMethodProxy>run:with:in: 593741548 OMenuMorph>drawOn: 593741456 CompiledMethod>valueWithReceiver:arguments: 593740508 >mareaHandleRun:with:in:forProxy: 593740416 AbstractCompiledMethodProxy>run:with:in: 593740324 Canvas>draw: 593740232 Canvas>drawMorph: 593740140 Morph>fullDrawOn: 593740028 FormCanvas>roundCornersOf:in:during: 593739936 Canvas>roundCornersOf:during: 593739808 Morph>fullDrawOn: 593739628 Canvas>fullDraw: 593739536 Canvas>fullDrawMorph: 593739004 WorldState>drawWorld:submorphs:invalidAreasOn: 593738104 Rectangle>allAreasOutsideList:startingAt:do: 593738012 Rectangle>allAreasOutsideList:do: 593738708 WorldState>drawWorld:submorphs:invalidAreasOn: 593737920 SequenceableCollection>do: 593737652 WorldState>drawWorld:submorphs:invalidAreasOn: 593737532 WorldState>displayWorld:submorphs: 593737440 FormCanvas>roundCornersOf:in:during: 593737348 Canvas>roundCornersOf:during: 593737244 WorldState>displayWorld:submorphs: 593737152 PasteUpMorph>privateOuterDisplayWorld 593737004 PasteUpMorph>displayWorld 593736912 WorldState>displayWorldSafely: 593736820 BlockClosure>on:do: 593736728 BlockClosure>ifError: 593736636 WorldState>displayWorldSafely: 593731064 WorldState>doOneCycleNowFor: 593730972 WorldState>doOneCycleFor: 593730880 PasteUpMorph>doOneCycle 564294164 >spawnNewProcess 564294036 BlockClosure>newProcess But this is just one example. Thanks in advance for any help Mariano

3 4

[patch] Don't strip scripts
by Aleksej Saushev 08 Dec '10

08 Dec '10

Hello! Stripping scripts isn't allowed across platforms. --- platforms/unix/config/make.cfg.in.orig 2010-12-07 21:00:03.000000000 +0300 +++ platforms/unix/config/make.cfg.in 2010-12-07 21:32:51.000000000 +0300 @@ -72,6 +72,7 @@ MKINSTALLDIRS= mkdir -p INSTALL= @INSTALL@ INSTALL_PROG= @INSTALL_PROGRAM@ $(INSTALL_ARGS) +INSTALL_SCRIPT= @INSTALL_SCRIPT@ $(INSTALL_ARGS) INSTALL_DATA= @INSTALL_DATA@ $(INSTALL_ARGS) UNINSTALL= $(SHELL) $(cfgdir)/uninstall AS= @AS@ --- platforms/unix/config/Makefile.install.orig 2010-12-07 21:00:03.000000000 +0300 +++ platforms/unix/config/Makefile.install 2010-12-07 21:24:27.000000000 +0300 @@ -19,8 +19,8 @@ $(MKINSTALLDIRS) $(ROOT)$(plgdir) $(MKINSTALLDIRS) $(ROOT)$(bindir) $(INSTALL_PROG) $(squeak) $(ROOT)$(plgdir) - $(INSTALL_PROG) $(squeaksh) $(ROOT)$(exec_prefix)/$(squeak) - $(INSTALL_PROG) $(binsqueaksh) $(ROOT)$(bindir)/$(squeak) + $(INSTALL_SCRIPT) $(squeaksh) $(ROOT)$(exec_prefix)/$(squeak) + $(INSTALL_SCRIPT) $(binsqueaksh) $(ROOT)$(bindir)/$(squeak) if test -n "$(VM_APP_ICONS)"; then \ $(INSTALL_DATA) $(ICONS) $(ROOT)$(VM_APP_ICONS); \ fi -- HE CE3OH...

2 2

Recursive not understood error encountered
by Mariano Martinez Peck 07 Dec '10

07 Dec '10

Hi. I am still playing with Objects as methods, and I am having yet another kind of crash :( This time, the image crash with the message "Recursive not understood error encountered". I am completly sure that I am not putting ANY proxy for any method of DNU, so I don't understand why I get this error. I copy the stacktrace. You can see 2 important things: 1) AbstractCompiledMethodProxy>run:with:in: 2) mareaHandleRun:with:in:forProxy: 1) is the proxy object and 2) is the handler. AbstractCompiledMethodProxy >> run: aSelector with: arguments in: aReceiver ^ CompiledMethodProxyHandler mareaHandleRun: aSelector with: arguments in: aReceiver forProxy: self CompiledMethodProxyHandler >> mareaHandleRun: aSelector with: arguments in: aReceiver forProxy: aProxy | aResult oldMethod | self mareaCheckLoop. self mareaLog: 'The method ', aSelector, ' is being executed with a receiver instance of ', aReceiver mareaClass name. oldMethod := self mareaUninstallForProxy: aProxy. aResult := oldMethod valueWithReceiver: aReceiver arguments: arguments. self mareaLog: 'The method ', aSelector, ' was executed with a receiver instance of ', aReceiver mareaClass name. ^aResult Does anyone has an idea of what can be happening? Thanks in advance, Recursive not understood error encountered 591830388 Rectangle>translateBy: 591830296 ECMenuMorph>prepareRectForNextRow: 591830116 CompiledMethod>valueWithReceiver:arguments: 591799776 >mareaHandleRun:with:in:forProxy: 591799684 AbstractCompiledMethodProxy>run:with:in: 591678220 OMenuMorph>drawOn: 591678128 CompiledMethod>valueWithReceiver:arguments: 591677672 >mareaHandleRun:with:in:forProxy: 591677580 AbstractCompiledMethodProxy>run:with:in: 591677488 Canvas>draw: 591677396 Canvas>drawMorph: 591677304 Morph>fullDrawOn: 591677192 FormCanvas>roundCornersOf:in:during: 591677100 Canvas>roundCornersOf:during: 591676972 Morph>fullDrawOn: 591676792 Canvas>fullDraw: 591676700 Canvas>fullDrawMorph: 591676204 WorldState>drawWorld:submorphs:invalidAreasOn: 591675472 Rectangle>allAreasOutsideList:startingAt:do: 591675380 Rectangle>allAreasOutsideList:do: 591675908 WorldState>drawWorld:submorphs:invalidAreasOn: 591675288 SequenceableCollection>do: 591675020 WorldState>drawWorld:submorphs:invalidAreasOn: 591674900 WorldState>displayWorld:submorphs: 591674808 FormCanvas>roundCornersOf:in:during: 591674716 Canvas>roundCornersOf:during: 591674612 WorldState>displayWorld:submorphs: 591674520 PasteUpMorph>privateOuterDisplayWorld 591674428 PasteUpMorph>displayWorld 591674336 WorldState>displayWorldSafely: 591674188 BlockClosure>on:do: 591674096 BlockClosure>ifError: 591674004 WorldState>displayWorldSafely: 591643764 WorldState>doOneCycleNowFor: 591643672 WorldState>doOneCycleFor: 591643580 PasteUpMorph>doOneCycle 560519040 >spawnNewProcess 560518912 BlockClosure>newProcess Debugger stopped. Program exited with status value:255.(gdb)

3 8

VM Maker: VMMaker-dtl.208.mcz
by squeak-dev-noreply＠lists.squeakfoundation.org 07 Dec '10

07 Dec '10

Dave Lewis uploaded a new version of VMMaker to project VM Maker: http://www.squeaksource.com/VMMaker/VMMaker-dtl.208.mcz ==================== Summary ==================== Name: VMMaker-dtl.208 Author: dtl Time: 7 December 2010, 7:48:42 am UUID: 5c1420f5-460e-47c5-8f70-2d180bee3a02 Ancestors: VMMaker-EstebanLorenzano.207 VMMaker 4.3.21 Translate ByteString>>findSubstring:in:startingAt:matchTable: in MiscPrimitivePlugin. Background: http://bugs.squeak.org/view.php?id=7479

1 0

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

Vm-dev December 2010