Compiling error in MS VC++ for kronrodintegral

[amar singh]

Hi,

 

I get the following compiling error in MS VC++ when trying to use KrondonIntegral class ( though it works with Borland C++). Could someone please give a hint?

\quantlib\quantlib-0.3.4\ql\math\kronrodintegral.hpp(52) : fatal error C1001: INTERNAL COMPILER ERROR
        (compiler file 'msc1.cpp', line 1794)
         Please choose the Technical Support command on the Visual C++
         Help menu, or open the Technical Support help file for more information

 

 

Thanks for your help,

Amar

---

Do you Yahoo!?
Protect your identity with Yahoo! Mail AddressGuard

[amar singh]

I think the problem is due to the signature of the function I pass in the operator(). I had earlier used as following :

 

double Fn(double x)

{

return x^2;
}

 

double val = KronrodIntegral(0.1,1000)(Fn,0,5);

 

This gave the above compiling error in MS VC++, while it works in Borland C++.

 

Writing it as

double val = KronrodIntegral(0.1,1000)(std::ptr_fun<double,double>(Fn),0,5);

 

solves the compiling error problem in MS VC++.

 

Don't know how it works though! (copied the usage from test-suite!)

 

Thanks all for your help,

Amar

amar singh <[hidden email]> wrote:

Hi,

 

I get the following compiling error in MS VC++ when trying to use KrondonIntegral class ( though it works with Borland C++). Could someone please give a hint?

\quantlib\quantlib-0.3.4\ql\math\kronrodintegral.hpp(52) : fatal error C1001: INTERNAL COMPILER ERROR
        (compiler file 'msc1.cpp', line 1794)
         Please choose the Technical Support command on the Visual C++
         Help menu, or open the Technical Support help file for more information

 

 

Thanks for your help,

Amar

---

Do you Yahoo!?
Protect your identity with Yahoo! Mail AddressGuard

---

Do you Yahoo!?
Protect your identity with Yahoo! Mail AddressGuard

[Luigi Ballabio-2]

On 2003.12.23 07:10, amar singh wrote:

> I think the problem is due to the signature of the function I pass in
> the operator(). I had earlier used as following :
>
> double Fn(double x)
> {
> return x^2;
> }
>
> double val = KronrodIntegral(0.1,1000)(Fn,0,5);
>
> This gave the above compiling error in MS VC++, while it works in
> Borland C++.
>
> Writing it as
> double val = KronrodIntegral(0.1,1000)(std::ptr_fun<double,double>
> (Fn),0,5);
>
> solves the compiling error problem in MS VC++.
>
> Don't know how it works though! (copied the usage from test-suite!)

Amar,
        the error you had wasn't an error in your code--it was an  
internal compiler error, meaning that VC++ got confused by something  
and could not even tell you why. In short, you triggered a bug in the  
compiler. The compiler documentation is not much help either--it merely  
suggests to rewrite your (legal) code in a different way. In short,  
they don't know why the compiler goes bang and hope that rewriting the  
code will cause the compiler to do something different.

Sorry if I cannot be of much more help,
                Luigi