Buffer Overrun Project, Problem Entering NULL in to stream

I’ve got a strange assignment from a class in which I’m supposed to initiate a buffer overrun to see how easy it is for someone to take advantage of the gets() function.

The basic program that I’m given is this:

Code:

typedef struct student
{
	char name[20];
	float amount;
	int other_stuff;
} person;

int _tmain(int argc, _TCHAR* argv[])
{
	person p; // Create a person account (on the stack)
	p.amount = 10.0; // set amount
	printf("enter name: ");
	gets(p.name); // get name (blindly ignoring buffer overrun)
	printf("Here is a $%6.2f gift for %s\n", p.amount, p.name);
	getch(); // Wait for user to type <Enter> before quitting
	return 0;
}

I understand that if I write more than 20 characters and the command prompt I will start writing in to the area in memory where “amount” is located so I can change it to anything that I want it to be. We are supposed to set the “amount” to 1,000,000 which I’ve figured out is – 0x00 0x24 0x74 0x49 when arranged in memory. I know that I can hold down alt and input the decimal equivalent of those numbers to overwrite memory but I can’t get a NULL to work for my first byte.

In a perfect world I would use this NULL after I enter my name then put in whatever I want up until I get to the amount area in memory and write in alt 000, alt 036, alt 116, alt 073. Unfortunately I can’t get alt 000 to register and I’m wondering if there is another way to put a NULL bit in to the stream at a command prompt.

Does anyone have any ideas?

***EDIT***

I might be wrong on what I need to enter as now looking in my debugger I see a different hex representation for 1,000,000 - 0x00 0x50 0xc3 0x47

gets() should automatcially add the null, when you hit ENTER.


1,000,000 decimal converts to F4240 hex as an integer. But, when it comes to storing a float, things get complicated, so, the debugger is probably the best way to do it.

Yeah but that's the problem, I want a NULL before I hit enter. In an ideal world I'd have a NULL after I enter my name so that it acts as a normal C string and then I would have a NULL in my representation of float 1,000,000 as well.

IE "Jo BobNULL--------------FLOAT VAL" would be my input at the command line where FLOAT VAL would be whatever characters are needed to make 1,000,000 in floating point arithmetic.