On 12-08-02 5:56 PM, Frank Shearar wrote:
> On 2 August 2012 20:50, Levente Uzonyi <leves at elte.hu> wrote:
> Maybe. I've got Chris Cunnington's de-Authored HudsonBuildTools, and
> now get a more interesting failure:
> http://squeakci.org/job/GitTest/9/console which shows a segfault
> during a CompilerNotifyingTest:
>
> Smalltalk stack dump:
> 0xbf8b9880 I Clipboard>clipboardText 2012164252: a(n) Clipboard
> 0xbf8b98a0 I Clipboard class>clipboardText 2017668068: a(n) Clipboard class
> 0xbf8b98c0 I SmalltalkEditor(Editor)>clipboardText 2034626872: a(n)
> SmalltalkEditor
> 0xbf8b98e8 I SmalltalkEditor(TextEditor)>cut 2034626872: a(n) SmalltalkEditor
> 0xbf8b9910 I CompilerNotifyingTest>enumerateAllSelections 2033658260:
> a(n) CompilerNotifyingTest
> 0xbf8b9928 M CompilerNotifyingTest>testATempShadowingAnotherTemp
> 2033658260: a(n) CompilerNotifyingTest
> 0xbf8b9940 M CompilerNotifyingTest(TestCase)>performTest 2033658260:
> a(n) CompilerNotifyingTest
> 0xbf8b9958 M [] in CompilerNotifyingTest(TestCase)>runCase 2033658260:
> a(n) CompilerNotifyingTest
> 0xbf8b9974 M BlockClosure>on:do: 2034618632: a(n) BlockClosure
> 0xbf8b999c M [] in CompilerNotifyingTest(TestCase)>timeout:after:
> 2033658260: a(n) CompilerNotifyingTest
> 0xbf8b99bc M BlockClosure>ensure: 2034619948: a(n) BlockClosure
> 0xbf8b99e4 M CompilerNotifyingTest(TestCase)>timeout:after:
> 2033658260: a(n) CompilerNotifyingTest
> 0xbf8b9a04 M [] in CompilerNotifyingTest(TestCase)>runCase 2033658260:
> a(n) CompilerNotifyingTest
> 0xbf8b9a24 M BlockClosure>ensure: 2034616652: a(n) BlockClosure
> 0xbf8b9a40 M CompilerNotifyingTest(TestCase)>runCase 2033658260: a(n)
> CompilerNotifyingTest
> 0xbf8b9a5c M [] in HDTestReport>runCase: 2033663528: a(n) HDTestReport
>
> I'll go look at VM switches tomorrow, unless you beat me to
> understanding the problem. I'll rerun the test and see if it's
> intermittent; those results will be at
> http://squeakci.org/job/GitTest/9/console
>
> frank
>

That's and interesting problem. I'm going to bet it's a freakish 
anomaly. Naturally, I was concerned about the HDHudsonBuild.st file I 
sent you. So I took an updated 4.4, ran it on a r2545, and loaded the 
file. I ran:

HDTestReport runPackages: #(
     'VersionNumberTests'
     )

and very satisfyingly it produces a VersionNumberTests-Test.xml file in 
the image directory. I remember spending a lot of my time away from 
Jenkins just testing that this part worked, that the HDHudsonBuild tools 
did just that. In production I believe it's the Emma plugin that moves 
such files to the place where they need to be to be seen on a Jenkins 
web page.
Thus I don't think the file I sent you can be the problem. I did the 
same thing with a r2559, which is the most up to date Cog vm for Mac. 
(You're running the most recent version, which is Linux only, I think - 
r2562 )
Perhaps you might try it, just for peace of mind. Load the file into an 
image on your desktop and execute #runPackages: as above to see the .xml 
files generated. Feeling confident in that part of the process makes 
facing Jenkins easier again. At least that was the way for me.
OK, I just broke out VirtualBox and loaded in the file into the 
4.3All-In-One and it cranked out a VersionNumberTests-Text.xml file in 
the Resources folder quite reliably.

So, in sum, I feel confident that the piece of code I gave you today is 
simple and reliable.

Chris