Can we use the tailcall optimization in this function If no

Can we use the tail-call optimization in this function? If no, explain why not. If yes, what is the difference in the number of executed instructions in f with and without the optimization?

}

Solution

}

Yes, we can use the tail call optimization in this function. Last line of function f is the tail call.

With any tail call, the function call itself can be optimized away and turned into what is effectively a goto. This means that the work to setup the stack before the function call and restore it afterwards can all be removed.

The unoptimized code might look something like this:

... ! executing inside f()

push EIP ! push current instruction pointer on stack

push a ! push variable \'a\' on the stack

push b ! push variable \'b\' on the stack

push c ! push variable \'c\' on the stack

push d ! push variable \'d\' on the stack

jmp g ! call g() by jumping to its address

... ! executing inside g()

pop g(a,b) ! prepare to return value of \'g(a,b)\'

pop c+d ! prepare to return value of \'c+d\'

pop EIP ! return to f()

pop d ! prepare to return value of \'d\'

pop c ! prepare to return value of \'c\'

pop b ! prepare to return value of \'b\'

pop a ! prepare to return value of \'a\'

pop EIP ! return to f() caller.

This shows the difference in number of executed instructions in f with and without the optimization.