Squeak-dev December 2014

squeak-dev@lists.squeakfoundation.org

35 participants
112 discussions

Spur with Immediate Floating Point Support implies a break
by Eliot Miranda 05 Dec '14

05 Dec '14

Hi All, some of you have been brave enough to use Spur and may have got used to being able to update. Recently I've updated Spur with support for immediate floating-point in 64-bit Spur. Alas these changes are not amenable to a straight-forward Monticello update. Now that I've updated Kernel.spur with these changes you'll not be able to simply update your image. There /may/ be a chance of being able to update if you first file-in MorphFloat.st (find attached). It worked for me. So in a recent SPur image, file-in MorphFloat.st and then update. If things get stuck on a partial update of Kernel.spur-eem.867(blah).mcd, then load Kernel.spur-eem.867.mcz manually and then update again. If this doesn't work apologies. What you can definitely do is upload the latest Spur image from www.mirandabanda.org/files/Cog/SpurImages/2014-12-01 and rebuild. -- best, Eliot

7 21

Re: [squeak-dev] Problems Running Squeak On FreeBSD 10.1
by John Briggs 05 Dec '14

05 Dec '14

On Fri, Dec 05, 2014 at 04:25:13AM +0000, B J wrote: > I recently installed FreeBSD 10.1 (Intel i386) on an external hard > drive, using Mate as the desktop and Slim as the login manager. I am > having problems in getting Squeak to run on that system. > > I used: > > pkg install squeak > > and the version that's installed is squeak-4.10.2_2 > > Running as root, I get the following messages: > > CHECKING cogvm > CHECKING squeakvm > Illegal instruction (core dumped) > > The result is a 1 GB file named squeakvm.core in the root directory. > > If I remove it and re-install it through as a port, I get the same > messages with squeakvm.core now located in: > > /usr/ports/lang/squeak/work/state/usr/local/lib/squeak/4.10.2-2614/ > > Squeak itself, without an image specified in the command line, appears > to run, so it seems that the installation has problems in reading > images. Try this method: create a local working directory (lwd) for Squeak. copy Squeak image and changes files to lwd. from the command line: cd lwd squeak squeak.image If this works create a shortcut to the desktop with the same commands for the shortcut > > Earlier, I tried different images and desktops, with the same result. > > By comparison, I installed it on an internal drive on the same machine > several months ago. I was using FreeBSD 10.0, Gnome 2 as the desktop, > and GDM as the login manager. I had no problems in running Squeak > with that configuration, though I believe the version is > squeak-4.10.2. > > I've made inquiries on the FreeBSD ports mailing list and in the > FreeBSD forum, but no solution so far. > > Did I miss something or do something wrong? Does anyone have any suggestions? > > Thank you. >

2 1

Re: netstyle.ch, Cog, systemd, and Socket>>initialize:
by Max Leske 03 Dec '14

03 Dec '14

> On 03.12.2014, at 19:22, Chris Cunnington <brasspen(a)gmail.com> wrote: > > >> On Dec 3, 2014, at 12:28 PM, Max Leske <maxleske(a)gmail.com <mailto:maxleske@gmail.com>> wrote: >> >> Hi Chris >> >>> On 03.12.2014, at 16:22, Chris Cunnington <brasspen(a)gmail.com <mailto:brasspen@gmail.com>> wrote: >>> >>> In July netstyle.ch <http://netstyle.ch/> said some exciting things about systemd and socket activation. [1] I have an observation and a question. >>> The observation is that to set Real Time Priority pursuant of the README for Cog you need to add the following in your Service or you will not be able to start coglinuxht. [2] >>> >>> [Service] >>> LimitRTPRIO=2 >>> >>> The question concerns the code in the systemd.cs file [4] which I don't think can work because it calls a non-existant primitive in the SocketPlugin. I suspect netstyle.ch <http://netstyle.ch/> wrote its patch for a very old virtual machine, one that still used #primitiveSocketCreate3Semaphores, which I don't think exists in contemporary SocketPlugin code. [3] In the image, there remain references to this non-existant primitive. They are employed by code in the systemd.cs file.[4] The culprit is Socket>>initialize: used in Socket class>>newSystemd. >> >> It seems you are right. We did use a Pharo1.3 image with an older VM to test the patch. However, I’m sure that we used the latest version from trunk for the patch (which was 3016 at the time). Also, if I’m not looking in the wrong place, that primitive is still there: http://squeakvm.org/cgi-bin/viewvc.cgi/squeak/trunk/src/plugins/SocketPlugi… <http://squeakvm.org/cgi-bin/viewvc.cgi/squeak/trunk/src/plugins/SocketPlugi…>. >> >> Does that resolve the issue? If not, please get back to me and I’ll try to fix the situation. > > Hi Max, > > Thank you very much for your quick response. I have no doubt that I've made a mistake and will look at it further. I've been at this alone for a while, so I'm sure your feedback will resolve the issue. If not, in about a week, I'd like to ask another question. Sure. > > Thanks to you and the netstyle.ch <http://netstyle.ch/> team for sharing this. It's a lot of fun. Thank you. It’s great that its something that people use and that doesn’t just collect virtual dust. > > Thanks, > Chris > >> >> Cheers, >> Max > > > >> >>> >>> So, Max, my question is, did you at netstyle.ch <http://netstyle.ch/> find a solution for this? >>> Does anybody have an idea of how to re-route Socket>>initialize: towards a primitive that actually exists? >>> >>> Thanks, >>> Chris >>> >>> >>> [1] http://lists.squeakfoundation.org/pipermail/squeak-dev/2014-July/179015.html <http://lists.squeakfoundation.org/pipermail/squeak-dev/2014-July/179015.html> >>> [2] http://www.freedesktop.org/software/systemd/man/systemd.exec.html <http://www.freedesktop.org/software/systemd/man/systemd.exec.html> >>> [3] http://squeakvm.org/cgi-bin/viewvc.cgi/squeak/trunk/platforms/unix/ <http://squeakvm.org/cgi-bin/viewvc.cgi/squeak/trunk/platforms/unix/> >>> [4] contents of the systemd.cs file [1]: >>> 'From Pharo1.3 of 16 June 2011 [Latest update: #13315] on 4 July 2014 at 3:20:20 pm'! >>> >>> !TcpListener methodsFor: '*systemd' stamp: 'nl 07/04/2014 15:17'! >>> haveSystemdAssignedPort >>> "LISTEN_FDS is set by systemd if the process/pharo-vm is started via port activation" >>> ^ (self primEnvironmentAtSymbol: 'LISTEN_FDS') >>> ifNotNilDo: [ :value | value asNumber = 1 ] >>> ifNil: [ false]! ! >>> >>> !TcpListener methodsFor: '*systemd' stamp: 'nl 07/04/2014 15:18'! >>> pvtNewListener: backlogSize >>> "Create a new socket that listens on our port. The backlog is how many simultaneous >>> connections to accept at the same time >>> >>> Use the socket assigned by systemd if: >>> - the command-line port number is equal to the portNumber to be opened >>> - there is an environment variable LISTEN_FDS with value 1 >>> Otherwise use normal port allocation. >>> " >>> | listener | >>> listener := (self haveSystemdAssignedPort >>> ifTrue: [ self socketClass newSystemd ] >>> ifFalse: [ self socketClass newTCP ]). >>> self socketsToDestroy add: listener. >>> self >>> bind: listener >>> withBacklogSize: backlogSize. >>> ^ listener! ! >>> >>> !TcpListener methodsFor: '*systemd' stamp: 'nl 04/29/2013 14:04'! >>> primEnvironmentAtSymbol: aSymbol >>> "Copy-paste from UnixOSProcessPlugin: Answer the value of an environment variable in the external OS process." >>> <primitive: 'primitiveEnvironmentAtSymbol' module: 'UnixOSProcessPlugin'> >>> ^ nil ! ! >>> >>> !Socket class methodsFor: '*systemd' stamp: 'nl 07/04/2014 15:23'! >>> newSystemd >>> "Create a socket assigned by systemd port activation" >>> | SystemdSocketType | >>> SystemdSocketType := 2. >>> self initializeNetwork. >>> ^[ super new initialize: SystemdSocketType ] >>> repeatWithGCIf: [ :socket | socket isValid not ]! ! >>> >> >

1 0

netstyle.ch, Cog, systemd, and Socket>>initialize:
by Chris Cunnington 03 Dec '14

03 Dec '14

In July netstyle.ch said some exciting things about systemd and socket activation. [1] I have an observation and a question. The observation is that to set Real Time Priority pursuant of the README for Cog you need to add the following in your Service or you will not be able to start coglinuxht. [2] [Service] LimitRTPRIO=2 The question concerns the code in the systemd.cs file [4] which I don't think can work because it calls a non-existant primitive in the SocketPlugin. I suspect netstyle.ch wrote its patch for a very old virtual machine, one that still used #primitiveSocketCreate3Semaphores, which I don't think exists in contemporary SocketPlugin code. [3] In the image, there remain references to this non-existant primitive. They are employed by code in the systemd.cs file.[4] The culprit is Socket>>initialize: used in Socket class>>newSystemd. So, Max, my question is, did you at netstyle.ch find a solution for this? Does anybody have an idea of how to re-route Socket>>initialize: towards a primitive that actually exists? Thanks, Chris [1] http://lists.squeakfoundation.org/pipermail/squeak-dev/2014-July/179015.html <http://lists.squeakfoundation.org/pipermail/squeak-dev/2014-July/179015.html> [2] http://www.freedesktop.org/software/systemd/man/systemd.exec.html <http://www.freedesktop.org/software/systemd/man/systemd.exec.html> [3] http://squeakvm.org/cgi-bin/viewvc.cgi/squeak/trunk/platforms/unix/ <http://squeakvm.org/cgi-bin/viewvc.cgi/squeak/trunk/platforms/unix/> [4] contents of the systemd.cs file [1]: 'From Pharo1.3 of 16 June 2011 [Latest update: #13315] on 4 July 2014 at 3:20:20 pm'! !TcpListener methodsFor: '*systemd' stamp: 'nl 07/04/2014 15:17'! haveSystemdAssignedPort "LISTEN_FDS is set by systemd if the process/pharo-vm is started via port activation" ^ (self primEnvironmentAtSymbol: 'LISTEN_FDS') ifNotNilDo: [ :value | value asNumber = 1 ] ifNil: [ false]! ! !TcpListener methodsFor: '*systemd' stamp: 'nl 07/04/2014 15:18'! pvtNewListener: backlogSize "Create a new socket that listens on our port. The backlog is how many simultaneous connections to accept at the same time Use the socket assigned by systemd if: - the command-line port number is equal to the portNumber to be opened - there is an environment variable LISTEN_FDS with value 1 Otherwise use normal port allocation. " | listener | listener := (self haveSystemdAssignedPort ifTrue: [ self socketClass newSystemd ] ifFalse: [ self socketClass newTCP ]). self socketsToDestroy add: listener. self bind: listener withBacklogSize: backlogSize. ^ listener! ! !TcpListener methodsFor: '*systemd' stamp: 'nl 04/29/2013 14:04'! primEnvironmentAtSymbol: aSymbol "Copy-paste from UnixOSProcessPlugin: Answer the value of an environment variable in the external OS process." <primitive: 'primitiveEnvironmentAtSymbol' module: 'UnixOSProcessPlugin'> ^ nil ! ! !Socket class methodsFor: '*systemd' stamp: 'nl 07/04/2014 15:23'! newSystemd "Create a socket assigned by systemd port activation" | SystemdSocketType | SystemdSocketType := 2. self initializeNetwork. ^[ super new initialize: SystemdSocketType ] repeatWithGCIf: [ :socket | socket isValid not ]! !

2 1

DeepCopier fix
by Stéphane Rollandin 02 Dec '14

02 Dec '14

Hello, I'm sending here a fix here for method DeepCopier>>fixDependents, which did not check that some weak associations may have a nil key. I have just got bitten by the bug, and the fix seemed straightforward. Now I did not investigate further, but maybe the test for nil keys should be moved to a specific implementation of #associationsDo: in WeakIdentityKeyDictionary. What do you think ? (fix attached) Stef

2 2

Re: [Pharo-dev] Fwd: [squeak-dev] IdentitySet>>collect:
by Eliot Miranda 02 Dec '14

02 Dec '14

On Tue, Dec 2, 2014 at 1:20 AM, stepharo <stepharo(a)free.fr> wrote: > What is the old behavior? > These are from Smalltalk-80 v2 of April 1, 1983 on 31 May 1983 at 9:10:35 am Object methods for private species "Answer the preferred class for reconstructing the receiver. For example, collections create new collections whenever enumeration messages such as collect: or select: are invoked. The new kind of collection is determined by the species of the original collection. Species and class are not always the same. For example, the species of Interval is Array." ^self class Collection methods for enumerating collect: aBlock "Evaluate aBlock with each of the receiver's elements as the argument. Collect the resulting values into a collection that is like the receiver. Answer the new collection. " | newCollection | newCollection _ self species new. self do: [:each | newCollection add: (aBlock value: each)]. ^newCollection select: aBlock "Evaluate aBlock with each of the receiver's elements as the argument. Collect into a new collection like the receiver, only those elements for which aBlock evaluates to true. Answer the new collection." | newCollection | newCollection _ self species new. self do: [:each | (aBlock value: each) ifTrue: [newCollection add: each]]. ^newCollection Interval methods for private species ^Array But there are also already uses for comparing: CharacterBlock methods for comparing = aCharacterBlock self species = aCharacterBlock species ifTrue: [^stringIndex = aCharacterBlock stringIndex] ifFalse: [^false] Interval methods for comparing = anInterval "Answer true if my species and anInterval species are equal, and if our starts, steps and sizes are equal." self species == anInterval species ifTrue: [^start = anInterval first and: [step = anInterval increment and: [self size = anInterval size]]] ifFalse: [^false] However, the species comment clearly states the usage pattern, even though it fails to mention the intent (to provide a mutable container with which to derive a new collection from the original). > In addition it would be good to have tests for such core library. > + 1 Stef > > Le 1/12/14 11:44, Sven Van Caekenberghe a écrit : > > > > Begin forwarded message: > > *From: *Bert Freudenberg <bert(a)freudenbergs.de> > *Subject: **Re: [squeak-dev] IdentitySet>>collect:* > *Date: *1 Dec 2014 11:35:00 CET > *To: *The general-purpose Squeak developers list < > squeak-dev(a)lists.squeakfoundation.org> > *Reply-To: *The general-purpose Squeak developers list < > squeak-dev(a)lists.squeakfoundation.org> > > On 01.12.2014, at 07:35, Colin Putney <colin(a)wiresong.com> wrote: > > > On Sun, Nov 30, 2014 at 6:20 AM, Bert Freudenberg <bert(a)freudenbergs.de> > wrote: > > >> How would you define “type” here? Same class? >> > > Yes, same class. It might be reasonable to answer instances of analogous > classes for weak collections, and expect the caller to use #collect:as: if > they want to retain weakness. > > Colin > > > That’s certainly simple: > > “collect: returns a collection of the same class as the receiver, > except for weak collections, which answer non-weak equivalents. To return a > collection of a different class, use collect:as:" > > - Bert - > > > The above definition makes perfect sense to me. The default should be > simple and predictable, the more complex and special cases can be handled > by a more convoluted API. > > And we should change callers that depend on the old behaviour. > > Sven > > > > > -- best, Eliot

1 0

Re: [Pharo-dev] Fwd: [squeak-dev] IdentitySet>>collect:
by Eliot Miranda 02 Dec '14

02 Dec '14

On Tue, Dec 2, 2014 at 1:20 AM, stepharo <stepharo(a)free.fr> wrote: > What is the old behavior? > These are from Smalltalk-80 v2 of Object methods for private species "Answer the preferred class for reconstructing the receiver. For example, collections create new collections whenever enumeration messages such as collect: or select: are invoked. The new kind of collection is determined by the species of the original collection. Species and class are not always the same. For example, the species of Interval is Array." ^self class Collection methods for enumerating collect: aBlock "Evaluate aBlock with each of the receiver's elements as the argument. Collect the resulting values into a collection that is like the receiver. Answer the new collection. " | newCollection | newCollection _ self species new. self do: [:each | newCollection add: (aBlock value: each)]. ^newCollection select: aBlock "Evaluate aBlock with each of the receiver's elements as the argument. Collect into a new collection like the receiver, only those elements for which aBlock evaluates to true. Answer the new collection." | newCollection | newCollection _ self species new. self do: [:each | (aBlock value: each) ifTrue: [newCollection add: each]]. ^newCollection Interval methods for private species ^Array But there are also already uses for comparing: CharacterBlock methods for comparing = aCharacterBlock self species = aCharacterBlock species ifTrue: [^stringIndex = aCharacterBlock stringIndex] ifFalse: [^false] Interval methods for comparing = anInterval "Answer true if my species and anInterval species are equal, and if our starts, steps and sizes are equal." self species == anInterval species ifTrue: [^start = anInterval first and: [step = anInterval increment and: [self size = anInterval size]]] ifFalse: [^false] However, the species comment clearly states the usage pattern, even though it fails to mention the intent (to provide a mutable container with which to derive a new collection from the original). In addition it would be good to have tests for such core library. > +1 > > Stef > > Le 1/12/14 11:44, Sven Van Caekenberghe a écrit : > > > > Begin forwarded message: > > *From: *Bert Freudenberg <bert(a)freudenbergs.de> > *Subject: **Re: [squeak-dev] IdentitySet>>collect:* > *Date: *1 Dec 2014 11:35:00 CET > *To: *The general-purpose Squeak developers list < > squeak-dev(a)lists.squeakfoundation.org> > *Reply-To: *The general-purpose Squeak developers list < > squeak-dev(a)lists.squeakfoundation.org> > > On 01.12.2014, at 07:35, Colin Putney <colin(a)wiresong.com> wrote: > > > On Sun, Nov 30, 2014 at 6:20 AM, Bert Freudenberg <bert(a)freudenbergs.de> > wrote: > > >> How would you define “type” here? Same class? >> > > Yes, same class. It might be reasonable to answer instances of analogous > classes for weak collections, and expect the caller to use #collect:as: if > they want to retain weakness. > > Colin > > > That’s certainly simple: > > “collect: returns a collection of the same class as the receiver, > except for weak collections, which answer non-weak equivalents. To return a > collection of a different class, use collect:as:" > > - Bert - > > > The above definition makes perfect sense to me. The default should be > simple and predictable, the more complex and special cases can be handled > by a more convoluted API. > > And we should change callers that depend on the old behaviour. > > Sven > > > > > -- best, Eliot

1 0

Daily Commit Log
by commits＠source.squeak.org 02 Dec '14

02 Dec '14

Changes to Trunk (http://source.squeak.org/trunk.html) in the last 24 hours: http://lists.squeakfoundation.org/pipermail/packages/2014-December/007385.h… Name: Kernel.spur-eem.887 Ancestors: Kernel-eem.887, Kernel.spur-eem.886 Kernel-eem.887 patched for Spur by SpurBootstrapMonticelloPackagePatcher Cog-eem.225 A few more cases in COntextPart where objectClass: should be used instead of class to avoid proxy issues. ============================================= http://lists.squeakfoundation.org/pipermail/packages/2014-December/007386.h… Name: Kernel-eem.888 Ancestors: Kernel-eem.887 Fix CompiledMethod class newFrom: ============================================= http://lists.squeakfoundation.org/pipermail/packages/2014-December/007387.h… Name: KernelTests-eem.283 Ancestors: KernelTests-eem.282 Fix methodClass test ============================================= http://lists.squeakfoundation.org/pipermail/packages/2014-December/007388.h… Name: Kernel-eem.889 Ancestors: Kernel-eem.888 Provide a plausible definition of CompiledMethod>>copyFrom: that at least works with compiled methods and e.g. Array. ============================================= http://lists.squeakfoundation.org/pipermail/packages/2014-December/007389.h… Name: Kernel.spur-eem.888 Ancestors: Kernel-eem.888, Kernel.spur-eem.887 Kernel-eem.888 patched for Spur by SpurBootstrapMonticelloPackagePatcher Cog-eem.226 Fix CompiledMethod class newFrom: ============================================= http://lists.squeakfoundation.org/pipermail/packages/2014-December/007390.h… Name: Kernel.spur-eem.889 Ancestors: Kernel-eem.889, Kernel.spur-eem.888 Kernel-eem.889 patched for Spur by SpurBootstrapMonticelloPackagePatcher Cog-eem.226 Provide a plausible definition of CompiledMethod>>copyFrom: that at least works with compiled methods and e.g. Array. =============================================

1 0

The Trunk: Kernel-eem.889.mcz
by commits＠source.squeak.org 01 Dec '14

01 Dec '14

Eliot Miranda uploaded a new version of Kernel to project The Trunk: http://source.squeak.org/trunk/Kernel-eem.889.mcz ==================== Summary ==================== Name: Kernel-eem.889 Author: eem Time: 1 December 2014, 1:09:35.682 pm UUID: 1fb0b0d6-b62e-4d04-a3d7-e72a25dafeae Ancestors: Kernel-eem.888 Provide a plausible definition of CompiledMethod>>copyFrom: that at least works with compiled methods and e.g. Array. =============== Diff against Kernel-eem.888 =============== Item was added: + ----- Method: CompiledMethod>>copyFrom: (in category 'copying') ----- + copyFrom: anotherObject + "Copy to myself all instance variables I have in common with anotherObject. + This is dangerous because it ignores an object's control over its own inst vars." + + <primitive: 168> + anotherObject isCompiledMethod + ifTrue: + [1 to: self numLiterals do: + [:i| self literalAt: i put: (anotherObject literalAt: i)]] + ifFalse: + [1 to: self numLiterals do: + [:i| self literalAt: i put: (anotherObject at: i)]]. + self initialPC to: (self basicSize min: anotherObject basicSize) do: + [:i| + self basicAt: i put: (anotherObject basicAt: i)]!

1 0

The Trunk: KernelTests-eem.283.mcz
by commits＠source.squeak.org 01 Dec '14

01 Dec '14

Eliot Miranda uploaded a new version of KernelTests to project The Trunk: http://source.squeak.org/trunk/KernelTests-eem.283.mcz ==================== Summary ==================== Name: KernelTests-eem.283 Author: eem Time: 1 December 2014, 1:08:10.406 pm UUID: ba98cc21-5dfb-4558-9cb5-8a6d58158ccd Ancestors: KernelTests-eem.282 Fix methodClass test =============== Diff against KernelTests-eem.282 =============== Item was changed: ----- Method: CompiledMethodTest>>testMethodClass (in category 'tests - accessing') ----- testMethodClass + | method cls binding | - | method cls | method := self class >> #returnTrue. self assert: #returnTrue equals: method selector. "now make an orphaned method by just deleting the class. + old: #unknown + pre-environment semantics: return Obsolete class + environment semantics: return binding's value, which will be nil" + (Smalltalk classNamed: #TUTU) ifNotNil: + [Smalltalk removeClassNamed: #TUTU]. - old: #unknown - new semantics: return Obsolete class" - Smalltalk removeClassNamed: #TUTU. cls := Object subclass: #TUTU instanceVariableNames: '' classVariableNames: '' poolDictionaries: '' category: 'KernelTests-Methods'. cls compile: 'foo ^ 10'. method := cls >> #foo. + binding := cls binding. + self assert: binding value equals: (Smalltalk classNamed: #TUTU) description: 'binding before class removal'. + self assert: (Smalltalk classNamed: #TUTU) equals: method methodClass description: 'methodClass before class removal'. Smalltalk removeClassNamed: #TUTU. + self assert: binding value equals: method methodClass description: 'methodClass after class removal'.! - self assert: cls equals: method methodClass description: 'methodClass after class removal'.!

1 0