Hi Chris,
I'm trying to run the tests for 1.4 in Squeak and I'm getting a failure in tryChangingIndex50 I'm using Squeak 4.4 with a recent Cog VM on Linux, and I loaded Magma 1.4 from Squeak Map.
The actual failure is Assertion Failed in the final assertion of tryChangingIndex50:
self should: [ err result commitConflicts first conflictingObjects anySatisfy: [ : each | each key = ma yourself ] ].
I'm assuming the tests pass for you? What VM and OS combination have you tried?
Regards, Stuart
I'm assuming the tests pass for you? What VM and OS combination have you tried?
Yes, they're passing for me. I've pasted my context from Help | About below.
Hmm, very strange. What OS you running?
Image ----- /home/cmm/Chris/development/Squeak/_test.image Squeak4.4 latest update: #12327 Current Change Set: Unnamed1 Image format 6505 (32 bit)
Monticello Working Copies ------------------------- 311Deprecated (nice.2) 39Deprecated (ar.19) Balloon (egp.19) BalloonTests (egp.2) BrpExtensions (cmm.11) * Collections (bf.496) Collections-BTree (lr.73) CollectionsTests (ul.194) * Compiler (eem.243, cmm.179) Compression (ul.31) EToys (bf.89) Environments (cwp.6) Exceptions (cmm.37) Files (cmm.119) FlexibleVocabularies (ar.12) GetText (nice.25) Graphics (bf.195) GraphicsTests (nice.30) Help-Squeak-Project (cmm.9) Help-Squeak-TerseGuide (dtl.2) HelpSystem-Core (ul.56) HelpSystem-Tests (ul.14) Installer-Core (bp.360) Kernel (nice.720) KernelTests (nice.238) Ma-Ascii-Report (cmm.9) Ma-Client-Server-Core (cmm.225) Ma-Client-Server-Squeak-Tester (cmm.1) Ma-Client-Server-Tester (cmm.157) Ma-Collections (cmm.141) Ma-Core (cmm.232) Ma-Installer-Core (cmm.33) Ma-Search (cmm.42) Ma-Serializer-Core (cmm.290) Ma-Serializer-Squeak-Core (cmm.2) Ma-Serializer-Tests (cmm.40) Ma-Squeak-Core (cmm.1) Ma-Statistics (cmm.28) Magma-Client (cmm.642) Magma-Server (cmm.455) Magma-Squeak-Client (cmm.6) Magma-Tester (cmm.406) Magma-Tools (cmm.68) Monticello (bf.527) MonticelloConfigurations (bf.107) Morphic (dtl.633) MorphicExtras (bf.104) MorphicTests (ar.18) Multilingual (ul.156) MultilingualTests (ul.15) Nebraska (ul.35) Network (dtl.138) NetworkTests (dtl.35) OSProcess (dtl.73) PackageInfo-Base (bf.64) PreferenceBrowser (bf.47) Protocols (ael.39) ReleaseBuilder (fbs.88) SMBase (dtl.131) SMLoader (cmm.77) SOLHashTables (cmm.16) ST80 (eem.139) SUnit (ul.89) SUnitGUI (eem.56) ScriptLoader (ul.333) Services-Base (ul.48) ShoutCore (cwp.36) Sound (ul.32) Squeak-Version (ar.4662) System (fbs.506) SystemChangeNotification-Tests (ul.19) SystemReporter (dtl.20) Tests (ar.176) ToolBuilder-Kernel (cwp.53) ToolBuilder-MVC (ul.29) ToolBuilder-Morphic (cwp.87) ToolBuilder-SUnit (cwp.17) Tools (eem.431) ToolsTests (fbs.56) Traits (ul.290) TraitsTests (ul.10) TrueType (ul.19) TrunkScript (cmm.5) Universes (nice.45) VersionNumber (cmm.4) VersionNumberTests (nice.3) WriteBarrier (cmm.44) XML-Parser (ael.35)
Virtual Machine --------------- /opt/4dst/thirdparty/squeak/lib/squeak/4.0-2640/squeak Croquet Closure Cog VM [CoInterpreter VMMaker.oscog-eem.240] Unix built on Dec 20 2012 13:53:54 Compiler: 4.1.2 20080704 (Red Hat 4.1.2-48) platform sources revision VM: r2640 http://www.squeakvm.org/svn/squeak/branches/Cog Plugins: r2545 http://squeakvm.org/svn/squeak/trunk/platforms/Cross/plugins CoInterpreter VMMaker.oscog-eem.240 uuid: f72e2cd1-8b0e-45af-acd7-06ba9bbc2a50 Dec 20 2012 StackToRegisterMappingCogit VMMaker.oscog-eem.240 uuid: f72e2cd1-8b0e-45af-acd7-06ba9bbc2a50 Dec 20 2012
It seems to be due to the VM - I was using the latest Pharo VM (from http://pharo.gforge.inria.fr/ci/vm/cog/linux/) . When I use the same VM as you, it works fine.
I hadn't realised that even the VMs had diverged (other than the NativeBoost stuff, but this was the regular Cog VM, not the NB one).
here's the details:
Virtual Machine --------------- /home/stuart/smalltalk/cog-linux-latest/CogVM Croquet Closure Cog VM [CoInterpreter VMMaker-oscog-EstebanLorenzano.166] Unix built on Dec 12 2012 16:08:11 Compiler: 4.6.3 platform sources revision git://gitorious.org/cogvm/blessed.git Commit: 452863bdfba2ba0b188e7b172e9bc597a2caa928 Date: 2012-12-07 16:49:46 +0100 By: Esteban Lorenzano estebanlm@gmail.com Jenkins build #5922 CoInterpreter VMMaker-oscog-EstebanLorenzano.166 uuid: 5773fcb9-2982-4507-8a9e-4308ec33731e Dec 12 2012 StackToRegisterMappingCogit VMMaker-oscog-EstebanLorenzano.166 uuid: 5773fcb9-2982-4507-8a9e-4308ec33731e Dec 12 2012
I wonder if it's just a case of a patch in the Squeak VM that hasn't made it to the Pharo one yet?
On 7 January 2013 06:13, Chris Muller asqueaker@gmail.com wrote:
I'm assuming the tests pass for you? What VM and OS combination have you tried?
Yes, they're passing for me. I've pasted my context from Help | About below.
Hmm, very strange. What OS you running?
Wow, that's strange! I can't think of anything in Magma that should be affected by the VM like that.
Maybe the Pharo VM is exposing a Magma problem that the standard Cog VM doesn't? I'd like to debug it to make sure. Where can I find the failing VM you're using?
On Sun, Jan 6, 2013 at 6:34 PM, Stuart Herring st-lists@stuartherring.com wrote:
It seems to be due to the VM - I was using the latest Pharo VM (from http://pharo.gforge.inria.fr/ci/vm/cog/linux/) . When I use the same VM as you, it works fine.
I hadn't realised that even the VMs had diverged (other than the NativeBoost stuff, but this was the regular Cog VM, not the NB one).
here's the details:
Virtual Machine
/home/stuart/smalltalk/cog-linux-latest/CogVM Croquet Closure Cog VM [CoInterpreter VMMaker-oscog-EstebanLorenzano.166] Unix built on Dec 12 2012 16:08:11 Compiler: 4.6.3 platform sources revision git://gitorious.org/cogvm/blessed.git Commit: 452863bdfba2ba0b188e7b172e9bc597a2caa928 Date: 2012-12-07 16:49:46 +0100 By: Esteban Lorenzano estebanlm@gmail.com Jenkins build #5922 CoInterpreter VMMaker-oscog-EstebanLorenzano.166 uuid: 5773fcb9-2982-4507-8a9e-4308ec33731e Dec 12 2012 StackToRegisterMappingCogit VMMaker-oscog-EstebanLorenzano.166 uuid: 5773fcb9-2982-4507-8a9e-4308ec33731e Dec 12 2012
I wonder if it's just a case of a patch in the Squeak VM that hasn't made it to the Pharo one yet?
On 7 January 2013 06:13, Chris Muller asqueaker@gmail.com wrote:
I'm assuming the tests pass for you? What VM and OS combination have you tried?
Yes, they're passing for me. I've pasted my context from Help | About below.
Hmm, very strange. What OS you running?
Magma mailing list Magma@lists.squeakfoundation.org http://lists.squeakfoundation.org/mailman/listinfo/magma
Here's the VM I was using: http://pharo.gforge.inria.fr/ci/vm/cog/linux/cog-linux-latest.zip
On 8 January 2013 02:27, Chris Muller asqueaker@gmail.com wrote:
Wow, that's strange! I can't think of anything in Magma that should be affected by the VM like that.
Maybe the Pharo VM is exposing a Magma problem that the standard Cog VM doesn't? I'd like to debug it to make sure. Where can I find the failing VM you're using?
Ok when I test with this VM it crashes just like Eliots older Cog VMs (the ones prior to 2640) in exactly the same spot -- in the #testSequenceNumberGenerator test.
The best way to work on Magma on Pharo is simply to run with the latest Cog VM's or with the "classic" (non-Cog) VM.
- Chris
On Mon, Jan 7, 2013 at 4:00 PM, Stuart Herring st-lists@stuartherring.com wrote:
Here's the VM I was using: http://pharo.gforge.inria.fr/ci/vm/cog/linux/cog-linux-latest.zip
On 8 January 2013 02:27, Chris Muller asqueaker@gmail.com wrote:
Wow, that's strange! I can't think of anything in Magma that should be affected by the VM like that.
Maybe the Pharo VM is exposing a Magma problem that the standard Cog VM doesn't? I'd like to debug it to make sure. Where can I find the failing VM you're using?
I figured that might be the case. I'll stick with Elliot's latest builds for now, but I'll have to revisit it once I start working on Pharo 2.0 compatibility, as the 2.0 images complain about the VM being too old (kind of funny considering the Pharo VMs appaear to be based on older Cog code in the first place).
At the moment I'm still working on getting through the tests on 1.3. Haven't had a whole lot of time this week, but I should be able to make some real progress next week.
On 10 January 2013 04:52, Chris Muller ma.chris.m@gmail.com wrote:
Ok when I test with this VM it crashes just like Eliots older Cog VMs (the ones prior to 2640) in exactly the same spot -- in the #testSequenceNumberGenerator test.
The best way to work on Magma on Pharo is simply to run with the latest Cog VM's or with the "classic" (non-Cog) VM.
- Chris
On Mon, Jan 7, 2013 at 4:00 PM, Stuart Herring st-lists@stuartherring.com wrote:
Here's the VM I was using: http://pharo.gforge.inria.fr/ci/vm/cog/linux/cog-linux-latest.zip
On 8 January 2013 02:27, Chris Muller asqueaker@gmail.com wrote:
Wow, that's strange! I can't think of anything in Magma that should be affected by the VM like that.
Maybe the Pharo VM is exposing a Magma problem that the standard Cog VM doesn't? I'd like to debug it to make sure. Where can I find the failing VM you're using?
magma@lists.squeakfoundation.org