Smalltalk › Squeak › Squeak VM

VM crashes on Alien tests on Linux.

_‹ Previous Topic Next Topic _›

Classic

List

Threaded

2 messages Options

laurent laffont

VM crashes on Alien tests on Linux.

Hi,

I've created ConfigurationOfVMMaker version 1.4 to build Alien plugins.

I use:

- PharoCore-1.0-10517

- VMMaker-dtl.170

- Alien-VMMaker-Plugins-John M McIntosh.8

- Alien-VMMaker-Support-John M McIntosh.8

- squeak vm rev. 2203

I've tried with NewsqueakIA32ABIPlugin and NewsqueakIA32ABIPluginAttic (which one should I use ?) as external.

My build script is here http://github.com/lolgzs/easy_squeakvm/blob/alien/easy_squeakvm.sh

The VM is successfully built. Then I load Alien tests using:

ConfigurationOfAlien loadTests.

When I run the tests in SUnit, the VM crashes. What should I check ?

Here's the output:

$ ./squeak ../PharoCore-1.0-10517/PharoCore-1.0-10517.image

SQUEAK_ENCODING=UTF-8

SQUEAK_PATHENC=UTF-8

SQUEAK_PLUGINS=/home/lol/sandbox/squeakvm/easysqueak/out/build/%n

+ exec /home/lol/sandbox/squeakvm/easysqueak/out/build/squeakvm ../PharoCore-1.0-10517/PharoCore-1.0-10517.image

4 characters came in as

c1 = (5)

c2 = ` (60)

c3 = (1)

c4 = q (71)

The 14 double are 0.000000 37731012097013058067317437078204362971852032595528259107423762531097905265514261096396025173302750110612706358114292371797115438181248956805323712176147123864545071420926082396075239300957542084816252981938270394157639412729055222913295109423106709185098753526806347776000.000000 1.000000 20.000000 30.000000 40.000000 50.000000 60.000000 70.000000 80.000000 90.000000 100.000000 110.000000 120.000000

The two floats are 0.000000 and inf

The 13 floats are 0.000000 0.000000 0.000000 0.000000 1.000000 20.000000 30.000000 40.000000 50.000000 60.000000 70.000000 80.000000 90.000000

The 14 floats are 0.000000 0.000000 0.000000 0.000000 1.000000 20.000000 30.000000 40.000000 50.000000 60.000000 70.000000 80.000000 90.000000 100.000000

The two floats are 0.000000 and 0.000000

The float 0.000000 and integer 159532624

The integer 4 and float 0.000000

The 7 floats are 0.000000 0.000000 0.000000 27965084.000000 1.000000 20.000000 30.000000

4 ints came in as

i1 = 5 (5)

i2 = 159532832 (9824720)

i3 = 4 (4)

i4 = 1 (1)

8 ints came in as

i1 = 9 (9)

i2 = 159532960 (98247a0)

i3 = 4 (4)

i4 = -1074779508 (bff02a8c)

i5 = 1 (1)

i6 = 20 (14)

i7 = 30 (1e)

i8 = 40 (28)

4 ints came in as

i1 = 3 (3)

i2 = 159533024 (98247e0)

i3 = 4 (4)

i4 = -1217854351 (b7690471)

4 shorts came in as

s1 = 5 (5)

s2 = 18496 (4840)

s3 = 2 (2)

s4 = 1 (1)

Segmentation fault

2019482440 AlienLibrary>ffiPrintString:

2019481780 AlienSunit>testCallingSquenceString

2019481688 TestCase>performTest

2019481596 TestCase>runCase

2019481504 BlockClosure>ensure:

2019481380 TestCase>runCase

2019481288 TestResult>runCase:

2019481172 BlockClosure>on:do:

2019481036 TestResult>runCase:

2019480944 BlockClosure>on:do:

2019480736 TestResult>runCase:

2019480644 TestCase>run:

2019480552 TestRunner>runTest:

2019480460 TestRunner>runSuite:

2019480276 SequenceableCollection>do:displayingProgress:

2019276848 SequenceableCollection>withIndexDo:

2019276732 SequenceableCollection>do:displayingProgress:

2019276640 ProgressInitiationException>defaultAction

2019276476 BlockClosure>ensure:

2019275852 ProgressInitiationException>defaultAction

2019275760 UndefinedObject>handleSignal:

2019275668 ContextPart>handleSignal:

2019275576 Exception>signal

2019275436 ProgressInitiationException>display:at:from:to:during:

2019275344 >display:at:from:to:during:

2019275232 String>displayProgressAt:from:to:during:

2019275140 SequenceableCollection>do:displayingProgress:

2019275004 TestRunner>basicRunSuite:do:

2019274912 BlockClosure>ensure:

2019274744 TestRunner>basicRunSuite:do:

2019274652 TestRunner>runSuite:

2019272372 TestRunner>runAll

2019272280 PluggableButtonMorph>performAction

2019272188 PluggableButtonMorphPlus>performAction

2019272096 PluggableButtonMorph>mouseUp:

2019272004 SequenceableCollection>do:

2019271212 PluggableButtonMorph>mouseUp:

2019271076 PluggableButtonMorphPlus>mouseUp:

2019270984 Morph>handleMouseUp:

2019270892 MouseButtonEvent>sentTo:

2019270800 Morph>handleEvent:

2019270708 Morph>handleFocusEvent:

2019270540 HandMorph>sendFocusEvent:to:clear:

2019270448 PasteUpMorph>becomeActiveDuring:

2019270356 BlockClosure>on:do:

2019270264 PasteUpMorph>becomeActiveDuring:

2019270148 HandMorph>sendFocusEvent:to:clear:

2019269988 HandMorph>sendEvent:focus:clear:

2019269896 HandMorph>sendMouseEvent:

2019269804 HandMorph>handleEvent:

2019269540 HandMorph>processEvents

2019269432 WorldState>doOneCycleNowFor:

2019269340 SequenceableCollection>do:

2019269248 WorldState>handsDo:

2019269156 WorldState>doOneCycleNowFor:

2019269028 WorldState>doOneCycleFor:

2019268936 PasteUpMorph>doOneCycle

1996578680 >?

1996578552 BlockClosure>?

Aborted

Cheers,

Laurent Laffont

David T. Lewis

Re: VM crashes on Alien tests on Linux.

On Mon, Apr 26, 2010 at 10:53:20PM +0200, laurent laffont wrote:

>
> Hi,
>
> I've created ConfigurationOfVMMaker version 1.4 to build Alien plugins.
>
> I use:
> - PharoCore-1.0-10517
> - VMMaker-dtl.170
> - Alien-VMMaker-Plugins-John M McIntosh.8
> - Alien-VMMaker-Support-John M McIntosh.8
> - squeak vm rev. 2203

Laurent,

Thanks for doing the ConfigurationOfVMMaker update.

> I've tried with NewsqueakIA32ABIPlugin and NewsqueakIA32ABIPluginAttic
> (which one should I use ?) as external.

I had the same question. John, is NewsqueakIA32ABIPluginAttic just
there to hold on to some no-longer-used primitives? It compiles fine
so I am assuming that it is safe to use (a class comment would be
good BTW).

> My build script is here
> http://github.com/lolgzs/easy_squeakvm/blob/alien/easy_squeakvm.sh
>
> The VM is successfully built. Then I load Alien tests using:
> ConfigurationOfAlien loadTests.

FYI: I made the updates to VMMaker platforms/Cross/vm, and can verify
that the plugin compiles cleanly (building a 32-bit plugin/VM on a
64-bit Linux platform). However, I did not load Alien and cannot verify
that the resulting plugin functions correctly.

Dave