[Blitz-support] Bus Error

[Thomas M. <tho...@gm...>]

Hi,

I appear to be getting the following bus error and I have tracked it  
down to my call of the Laplacian2D(P) method but I am not sure what I  
am doing wrong. The error that I am getting is

Process:         a.out [5280]
Path:            ./a.out
Identifier:      a.out
Version:         ??? (???)
Code Type:       X86 (Native)
Parent Process:  bash [2437]

Date/Time:       2009-03-06 12:54:46.466 -0600
OS Version:      Mac OS X 10.5.6 (9G55)
Report Version:  6

Exception Type:  EXC_BAD_ACCESS (SIGBUS)
Exception Codes: KERN_PROTECTION_FAILURE at 0x0000000000031800
Crashed Thread:  0

Thread 0 Crashed:
0   a.out                         	0x00006a85  
blitz::FastArrayIterator<double, 2>::shift(int, int) + 43
1   a.out                         	0x00006ae6  
blitz::FastArrayIterator<double, 2>::T_numtype  
blitz::Laplacian2D<blitz::FastArrayIterator<double, 2>  
 >(blitz::FastArrayIterator<double, 2>&) + 78
2   a.out                         	0x00006bbf  
blitz::Laplacian2D_et<double, 2>::fastRead(int) + 55
3   a.out                         	0x00006bfc  
blitz::_bz_ArrayExpr<blitz::Laplacian2D_et<double, 2> >::fastRead(int)  
+ 24
4   a.out                         	0x0000c565 blitz::Array<double, 2>&  
blitz::Array<double,  
2 
 > 
::evaluateWithStackTraversalN 
<blitz::_bz_ArrayExpr<blitz::Laplacian2D_et<double, 2> >,  
blitz::_bz_update<double, double>  
 >(blitz::_bz_ArrayExpr<blitz::Laplacian2D_et<double, 2> >,  
blitz::_bz_update<double, double>) + 787
5   a.out                         	0x0000c85d blitz::Array<double, 2>&  
blitz::Array<double,  
2>::evaluate<blitz::_bz_ArrayExpr<blitz::Laplacian2D_et<double, 2> >,  
blitz::_bz_update<double, double>  
 >(blitz::_bz_ArrayExpr<blitz::Laplacian2D_et<double, 2> >,  
blitz::_bz_update<double, double>) + 81
6   a.out                         	0x0000c8aa blitz::Array<double, 2>&  
blitz::Array<double,  
2>::operator=<blitz::_bz_ArrayExpr<blitz::Laplacian2D_et<double, 2> >  
 >(blitz::ETBase<blitz::_bz_ArrayExpr<blitz::Laplacian2D_et<double, 2>  
 > > const&) + 58
7   a.out                         	0x000030cd  
init(blitz::Array<double, 2>&, blitz::Array<double, 2>&,  
blitz::Array<double, 3>&, blitz::Array<double, 3>&) + 939
8   a.out                         	0x000036a0 main + 592
9   a.out                         	0x0000264e start + 54

I'm not sure if any of this is useful but who knows.

The offending code appears to be
void init(Array<double,2>& c, Array<double,2>& P, Array<double,3>&  
pStep, Array<double,3>& pPower){
	//Defines a 1m^3 domain
	cout << "e" << endl;
	Array<double,3> pDiff(2*n+1,N,N);
	Array<double,2> temp(N,N);
	cout << "f" << endl;
	
	pDiff = 0;
	temp = 0;
	
	spatialStep = 1.0/(N-1);
	
	timeStep = .1;
	//Sets the conductive velocity
	c = 0.2;
	//Initial conditions
	using namespace blitz::tensor;
	int cr = N/2-1;
	int cc = 7.0*N/8.0-1.0;
	float s2 = 64.0*9.0/pow2(N/2);
	P = exp(-(pow2(i-cr) * s2)); //gaussian pulse
	cout << "g" << endl;

	temp = Laplacian2D(P);
	cout << "h" << endl;
	pDiff(0, Range::all(), Range::all()) = P;
	cout << "i" << endl;
	pDiff(1, Range::all(), Range::all()) = temp;
	cout << "j" << endl;
	
	for(int i = 2; i < n; i++) //Finds the Nth derivitave.  
Laplacian2D(P), etc
	{	
		temp = pDiff(i-1,Range::all(),Range::all());
		cout << "k" << endl;
		pDiff(i,Range::all(), Range::all()) = Laplacian2D(temp);
	}
	
	
}

Where the 	temp = Laplacian2D(P); is throwing the error.

	Array<double, 3> pStep(2*n+1,N,N), pPower(2*n+1,N,N); are defined as  
such.