Here's the top of the stack when this file plugin problem happens. It really looks like some changes have occurred in the image and the Simulator hasn't cought up yet. For instance using ByteSymbol instead of Symbol. I'll do what I can to bring this current.
12814BC StandardFileStream>>#basicNext args: temps: 04 stack: 1281394 UTF8TextConverter>>#nextFromStream: args: 1280688 temps: 04 04 04 04 04 04 04 04 04 stack: 12811EC MultiByteFileStream>>#next args: temps: 04 04 04 stack: 1281190 MultiByteFileStream>>#next: args: 191 temps: 12813F0 03 04 stack: 1268024 FileDirectory class>>#openSources:andChanges:forImage: args: 12687A8 1268BF4 10FB194 temps: 127B3B0 1280688 CA19FC CA1A78 stack:
----- Original Message ----- From: "Rob Withers" email@example.com To: firstname.lastname@example.org; "Squeak Virtual Machine Development Discussion" email@example.com Sent: Wednesday, October 24, 2007 10:17 PM Subject: Re: [Vm-dev] Any patch for InterpreterSimulator
Thanks, John. I fixed this in ObjectMemory>>#initializeObjectMemory:, since I figured it belonged there more than in the Interpreter. On running further, I also had to add fixes to the following, all switching from Smalltalk to SmalltalkImage current: InterpreterSimulator>>#primitiveGetAttribute InterpreterSimulator>>#vmPathGet:Length: InterpreterSimulator>>#vmPathSize
Then I hit FilePluginSimulator>>#primitiveFileRead. This gets its file from: file := self fileValueOf: (interpreterProxy stackValue: 3). where the stackValue is 20, which is not an integer object and the file is nil. It's beyond me.
How can I report these to Mantis. Is there a website for it?
----- Original Message ----- From: "John M McIntosh" firstname.lastname@example.org To: "Squeak Virtual Machine Development Discussion" email@example.com Sent: Wednesday, October 24, 2007 7:20 PM Subject: Re: [Vm-dev] Any patch for InterpreterSimulator
- primitiveSetGCSemaphore: Indicates a semaphore (index) to be
signaled whenever a garbage collection occurs.
it should have a value of zero if it's not used. Obviously in C that's zero by the magic of auto-initialization of that C variable, and so the code runs without care, but in Smalltalk it's nil and you die.
I'd make a mantis report and supply a patch to initialize the value explicitly in the interpreter startup code initializeInterpreter: so that the simulator will run.
On Oct 24, 2007, at 7:09 PM, Rob Withers wrote:
My first post on this list, and it may not be my last. I am trying to run the InterpreterSimulator adn am having issues. Mathieu Suen mentioned this list and that there may be a patch for it. Does anyone know of a patch?
My current issue is: In postGCAction, the gcSemaphoreIndex is nil, which blows up in signalSemaphoreWithIndex:. Trying to be a bit more self sufficient, I put a halt in primitiveSetGCSemaphore and reran it, but it never breaks, so it never seems to get set. Full stop.
thank you, Rob
=== John M. McIntosh firstname.lastname@example.org Corporate Smalltalk Consulting Ltd. http://www.smalltalkconsulting.com ======================================================================== ===