CS 536Final ExamSaturday, May 17, 20087:45 AM- 9:45 AMAnswer any four questions. (If you answer more, only the first four will count.) Each question is worth 25 points. Please try to...1. Assume that we add a conditional expression of the form (Expr1 ? Expr2 : Expr3) to CSX. Expr1 is an expression that returns a...2. (a) (20 points) In CSX we type check a call to method M by first type checking M’s declaration. Then the actual parameters in...3. Assume we add a new operator to CSX, the delay operator. The expression delay expr returns a suspended execution of expr (ess...4. Assume we are translating a function or procedure call in CSX. Explain how we could determine the exact number of stack locat...5. A common compiler optimization is inlining. A call to a method M is inlined when the call is replaced with the body of M, with formal parameters initialized to the actual parameters used in the call.6. (a) (10 points) Code for an if statement is generated on the assumption that the value of the control expression will not be ...CS 536Final ExamSaturday, May 17, 20087:45 AM— 9:45 AM1325 Computer ScienceInstructionsAnswer any four questions. (If you answer more, only the first four will count.) Each question is worth 25 points. Please try to make your answers neat and coherent. Remember, if we can’t read it, it’s wrong. Partial credit will be given, so try to put something down for each question (a blank answer always gets 0 points!).1. Assume that we add a conditional expression of the form (Expr1 ? Expr2 : Expr3) to CSX. Expr1 is an expression that returns a bool. If Expr1 is true, expression Expr2 is eval-uated, and its value is the value of the conditional expression. If Expr1 is false, expres-sion Expr3 is evaluated, and its value is the value of the conditional expression. Outline the changes that would be needed in a CSX type-checker and code generator to handle conditional expressions. Illustrate your answer using the following example: a = (i != 0 ? j/i : 0); 2. (a) (20 points) In CSX we type check a call to method M by first type checking M’s declaration. Then the actual parameters in the call to M are type checked. Finally, the number, type and kind of parameters found in the call are compared with the number, type and kind of parameters specified in M’s declaration. Consider the following alternative. In a declaration of a method P, no types are given to P’s parameters; they are simply given names. For example, int P(a,b,c) { ... } When a call to P is found, the parameters in the call are type checked (as usual) and then these types are used as definitions of the types of P’s parameters. What changes are needed in your type checking of method declarations and calls to implement this change?(b) (5 points) A potential difficulty in the approach suggested in part (a) is that different calls to Pmay differ in the types used for a particular parameter. How would you handle a non-unique type for a parameter in different calls to P?-2-3. Assume we add a new operator to CSX, the delay operator. The expression delay expr returns a suspended execution of expr (essentially a function of zero arguments whose body is return expr). The operation force suspension forces the given suspension (created by a delay operation) to be evaluated, returning the value of the suspended expression. For example, s=delay delete_file(f); creates a suspended call to delete_file with parameter f. The delete_file oper-ation is not done yet, it is just “set up.” At some later time the suspension created by the delay may be activated by using the force operation. Then the delete_fileoperation is performed. For example, r=force s; forces the suspended delete_file(f) operation to run, returning its result to r. Explain how you would implement delay and force in CSX. How can you be sure that all the variables visible when the delay was executed are still accessible when the force is executed? 4. Assume we are translating a function or procedure call in CSX. Explain how we could determine the exact number of stack locations that will be needed to evaluate and store (on the stack) the parameters for the call. Illustrate your technique on the following call: p(a+g(1,2), x-(y-(p+1))); 5. A common compiler optimization is inlining. A call to a method M is inlined when the call is replaced with the body of M, with formal parameters initialized to the actual parameters used in the call.What are the advantages of inlining? What are the disadvantages of inlining? Under what circumstances should inlining not be performed?Given a CSX program in AST form, how would you inline a a particular call of some method M? 6. (a) (10 points) Code for an if statement is generated on the assumption that the value of the control expression will not be known until run-time. But in some cases the value of the control expression is known at compile-time. The simplest such case if when the control expression is just the boolean literal true or false. What changes would you make in your CSX code generator for if statements to handle the special case of a control expression that is either the literal true or the literal false?-3-(b) (5 points) The special case handled in part (a) is uncommon. More common is the case in which the control expression is an identifier declared to be a boolean constant with a literal initializer. For example, const debug = true; ... if (debug) ... What changes are needed to your solution to part (a) to include the case of identifiers declared to be boolean constants? (c) (10 points) It may occur that the control expression of an if statement contains boolean opera-tors (&&, ||, !) whose operands are all boolean literals or boolean constants with lit-eral initializers. For example, const debug1 = true; const debug2 = false; ... if (debug1 || debug2) ... What changes are needed to your solution to part (b) to include the case of boolean operators whose operands are all boolean literals or boolean