Seg fault when port is bound to dummay signal in before_end_of_elaboration

[rahuljn]

Hello

 

In the following example I am tying to connect the unconnected port to dummy channel in before_end_of_elaboration callback. But it gives me seg fault

 

#include "systemc.h"

class test : public sc_module {
    public:
    sc_in<bool> in1;
    SC_HAS_PROCESS(test);
    test(sc_module_name name){
        SC_THREAD(fun);
    }

    void fun(){
    }

    void before_end_of_elaboration(){
        sc_signal<bool> sig;
        in1.bind(sig);
    }
};

int sc_main(int, char**){
    test t("t");
    sc_start();
    return 0;
}

 

Can you please help.

 

Also the standard says that @page 122 that the time at which deferred port binding is complete is implementation defined, so where I shoudl check my unconnected ports in the module"

 

Thanks

RahulJn

[Roman Popov]

Well, in my case your simulation does not produces segfault. But obviously you have a problem your code:

 

The problem is that you create your signal on call stack.  Basically it is local to before_end_of_elaboration function.

Once control returns from before_end_of_elaboration, signal sig is destroyed. 

 

So if you will try to read from port in1 you will have a segfault:

void fun(){
in1.read();
}

Solution is to make your signal a member of module:

class test : public sc_module {
public:
sc_in<bool> in1;
sc_signal<bool> sig;

SC_HAS_PROCESS(test);
test(sc_module_name name){
SC_THREAD(fun);
}

void fun(){
in1.read();
}

void before_end_of_elaboration(){
in1.bind(sig);
}
};

[rahuljn]

Hello

 

Thanks for your comments. But in my case I am not interested in reading that port value during the simulation as I am explicitly stubing that.

Can you confirm you are not getting seg fault with my example(without any change) ?

 

In my env I am using Ubuntu, gcc 4.8.4 and SystemC 2.3.1 and I am getting the seg fault.

 

 

 

Thanks

RJ

[Roman Popov]

I've checked on Windows now,

yes it's segfaults when tries to dereference pointer to sig during deferred biding. Unfortunately there is no mechanism in C++ to check if pointer is valid, so SystemC kernel can't save us from this type of mistakes.