Jump to content

  • Log In with Google      Sign In   
  • Create Account

VBO and glIsBuffer


Old topic!
Guest, the last post of this topic is over 60 days old and at this point you may not reply in this topic. If you wish to continue this conversation start a new topic.

  • You cannot reply to this topic
4 replies to this topic

#1 The Magical Pot   Members   -  Reputation: 178

Like
0Likes
Like

Posted 07 July 2014 - 06:59 AM

Hi!

 

I'm working with VBOs and I suddenly sumbled across a very weird error. My compiler is giving me a run-time error saying that this statement can't be executed:

if ( glIsBuffer( VertexBuffer ) )

This statement can be found in this destructor:

VBOQuad::~VBOQuad()
{
	//Clear vertex buffer
	if ( glIsBuffer( VertexBuffer ) )
		glDeleteBuffers( 1, &VertexBuffer );

	//Clear index buffer
	if ( glIsBuffer( IndexBuffer ) )
		glDeleteBuffers( 1, &IndexBuffer );

	VertexBuffer = NULL;
	IndexBuffer = NULL;
}

I've tried to track this error down to find some sort of context, but I've found that the run-time error emerges before the first line of the main function is executed. The description of glIsBuffer says that the function will return true if the argument is the name of an existing buffer, and it will return false if the argument is zero or a non-zero value. The debugger says:

 

VertexBuffer CXX0030: Error: expression cannot be evaluated
IndexBuffer CXX0030: Error: expression cannot be evaluated
 
 
This would, to me, indicate that VertexBuffer doesn't exist in memory, which in turn would mean that VertexBuffer is basically NULL, which is defined as zero, so there souldn't be a problem. That's probably not right since it's crashing at that statement.
 
The main problem I'm having is that the application crashes before the main function is executed, so for all I know, the destructor hasn't even been called yet, but it's trying to execute it, by the looks of it. My question is what the problem usually is in a situation like this where the application crashes before the main function is executed.
 
Thanks!
 
EDIT: I have added a screen dump of the Stack Trace.

Attached Thumbnails

  • Call Stack.jpg

Edited by The Magical Pot, 07 July 2014 - 11:07 AM.


Sponsor:

#2 Wh0p   Members   -  Reputation: 340

Like
2Likes
Like

Posted 07 July 2014 - 09:05 AM

Before the program enters the entry point of main() global variables (like 'quad' in the example) are allocated on the stack.

VBOQuad quad;
int main () { ... }

I can't say much to the error but have you checked the stack trace of you application when it crashed? This might give a clue what functions are called that lead to the destruction of this object.


Edited by Wh0p, 07 July 2014 - 09:05 AM.


#3 Sponji   Members   -  Reputation: 1257

Like
2Likes
Like

Posted 07 July 2014 - 09:08 AM

Sounds like there's no opengl context active when you try to call those opengl functions?


Derp

#4 Wh0p   Members   -  Reputation: 340

Like
0Likes
Like

Posted 07 July 2014 - 12:08 PM

So from the stack trace you have posted the error might found in the vector<Button>.

 

Looks like the class Button uses VBOQuad, so if there is somewhere a vector<Button> floating in your code the destructor of VBOQuad might be called while the vector is resized.

 

I don't know what State_1::State_1() is or does (static function???) but looks like that function is called and the error is caused from here by push_back which triggers the resize of the vector. I quess Sponji had the right hunch there. Also check that you are not double deleting any objects.

 

Good luck :)



#5 The Magical Pot   Members   -  Reputation: 178

Like
0Likes
Like

Posted 07 July 2014 - 12:12 PM

I managed to fix this at last. The call stack said that the error occurred after main was called and I realized that I used an incorrect method at first to check where the application crashed. When the application was to delete an object of type VBOQuad, I checked if the object existed in memory by comparing it to NULL, but I tried to print out the value of an uninitialized pointer of this type and I realized that such a pointer is not necessarily equal to NULL. In the end, an uninitialized pointer was seen as an initialized one, thus the application would try to delete an object that didn't exist in memory.

 

Thanks for your help! :)






Old topic!
Guest, the last post of this topic is over 60 days old and at this point you may not reply in this topic. If you wish to continue this conversation start a new topic.



PARTNERS