Library Smallstep

Smallstep: Small-step Operational Semantics


Require Export Imp.

The evaluators we have seen so far (e.g., the ones for aexps, bexps, and commands) have been formulated in a "big-step" style -- they specify how a given expression can be evaluated to its final value (or a command plus a store to a final store) "all in one big step."
This style is simple and natural for many purposes -- indeed, Gilles Kahn, who popularized its use, called it natural semantics. But there are some things it does not do well. In particular, it does not give us a natural way of talking about concurrent programming languages, where the "semantics" of a program -- i.e., the essence of how it behaves -- is not just which input states get mapped to which output states, but also includes the intermediate states that it passes through along the way, since these states can also be observed by concurrently executing code.
Another shortcoming of the big-step style is more technical, but critical in some situations. To see the issue, suppose we wanted to define a variant of Imp where variables could hold either numbers or lists of numbers (see the HoareList chapter for details). In the syntax of this extended language, it will be possible to write strange expressions like 2 + nil, and our semantics for arithmetic expressions will then need to say something about how such expressions behave. One possibility (explored in the HoareList chapter) is to maintain the convention that every arithmetic expressions evaluates to some number by choosing some way of viewing a list as a number -- e.g., by specifying that a list should be interpreted as 0 when it occurs in a context expecting a number. But this is really a bit of a hack.
A much more natural approach is simply to say that the behavior of an expression like 2+nil is undefined -- it doesn't evaluate to any result at all. And we can easily do this: we just have to formulate aeval and beval as Inductive propositions rather than Fixpoints, so that we can make them partial functions instead of total ones.
However, now we encounter a serious deficiency. In this language, a command might fail to map a given starting state to any ending state for two quite different reasons: either because the execution gets into an infinite loop or because, at some point, the program tries to do an operation that makes no sense, such as adding a number to a list, and none of the evaluation rules can be applied.
These two outcomes -- nontermination vs. getting stuck in an erroneous configuration -- are quite different. In particular, we want to allow the first (permitting the possibility of infinite loops is the price we pay for the convenience of programming with general looping constructs like while) but prevent the second (which is just wrong), for example by adding some form of typechecking to the language. Indeed, this will be a major topic for the rest of the course. As a first step, we need a different way of presenting the semantics that allows us to distinguish nontermination from erroneous "stuck states."
So, for lots of reasons, we'd like to have a finer-grained way of defining and reasoning about program behaviors. This is the topic of the present chapter. We replace the "big-step" eval relation with a "small-step" relation that specifies, for a given program, how the "atomic steps" of computation are performed.

A Toy Language

To save space in the discussion, let's go back to an incredibly simple language containing just constants and addition. (We use single letters -- C and P -- for the constructor names, for brevity.) At the end of the chapter, we'll see how to apply the same techniques to the full Imp language.

Inductive tm : Type :=
  | C : nattm
  | P : tmtmtm.
Tactic Notation "tm_cases" tactic(first) ident(c) :=
  first;
  [ Case_aux c "C" | Case_aux c "P" ].

Here is a standard evaluator for this language, written in the same (big-step) style as we've been using up to this point.

Fixpoint evalF (t : tm) : nat :=
  match t with
  | C nn
  | P a1 a2evalF a1 + evalF a2
  end.

Now, here is the same evaluator, written in exactly the same style, but formulated as an inductively defined relation. Again, we use the notation t || n for "t evaluates to n."
(E_Const) C n || n
t1 || n1 t2 || n2
(E_Plus) P t1 t2 || C (n1 + n2)

Reserved Notation " t '||' n " (at level 50, left associativity).

Inductive eval : tmnatProp :=
  | E_Const : n,
      C n || n
  | E_Plus : t1 t2 n1 n2,
      t1 || n1
      t2 || n2
      P t1 t2 || (n1 + n2)

  where " t '||' n " := (eval t n).

Tactic Notation "eval_cases" tactic(first) ident(c) :=
  first;
  [ Case_aux c "E_Const" | Case_aux c "E_Plus" ].

Module SimpleArith1.

Now, here is a small-step version.
(ST_PlusConstConst) P (C n1) (C n2) ==> C (n1 + n2)
t1 ==> t1'
(ST_Plus1) P t1 t2 ==> P t1' t2
t2 ==> t2'
(ST_Plus2) P (C n1) t2 ==> P (C n1) t2'

Reserved Notation " t '==>' t' " (at level 40).

Inductive step : tmtmProp :=
  | ST_PlusConstConst : n1 n2,
      P (C n1) (C n2) ==> C (n1 + n2)
  | ST_Plus1 : t1 t1' t2,
      t1 ==> t1'
      P t1 t2 ==> P t1' t2
  | ST_Plus2 : n1 t2 t2',
      t2 ==> t2'
      P (C n1) t2 ==> P (C n1) t2'

  where " t '==>' t' " := (step t t').

Tactic Notation "step_cases" tactic(first) ident(c) :=
  first;
  [ Case_aux c "ST_PlusConstConst"
  | Case_aux c "ST_Plus1" | Case_aux c "ST_Plus2" ].

Things to notice:
  • We are defining just a single reduction step, in which one P node is replaced by its value.
  • Each step finds the leftmost P node that is ready to go (both of its operands are constants) and rewrites it in place. The first rule tells how to rewrite this P node itself; the other two rules tell how to find it.
  • A term that is just a constant cannot take a step.
Let's pause and check a couple of examples of reasoning with the step relation...
If t1 can take a step to t1', then P t1 t2 steps to P t1' t2:

Example test_step_1 :
      P
        (P (C 0) (C 3))
        (P (C 2) (C 4))
      ==>
      P
        (C (0 + 3))
        (P (C 2) (C 4)).
Proof.
  apply ST_Plus1. apply ST_PlusConstConst. Qed.

Exercise: 1 star (test_step_2)

Right-hand sides of sums can take a step only when the left-hand side is finished: if t2 can take a step to t2', then P (C n) t2 steps to P (C n) t2':

Example test_step_2 :
      P
        (C 0)
        (P
          (C 2)
          (P (C 0) (C 3)))
      ==>
      P
        (C 0)
        (P
          (C 2)
          (C (0 + 3))).
Proof.
Admitted.

Relations

We will be using several different step relations, so it is helpful to generalize a bit...
A (binary) relation on a set X is a family of propositions parameterized by two elements of X -- i.e., a proposition about pairs of elements of X.

Definition relation (X: Type) := XXProp.

Our main examples of such relations in this chapter will be the single-step and multi-step reduction relations on terms, ==> and ==>*, but there are many other examples -- some that come to mind are the "equals," "less than," "less than or equal to," and "is the square of" relations on numbers, and the "prefix of" relation on lists and strings.
One simple property of the ==> relation is that, like the evaluation relation for our language of Imp programs, it is deterministic.
Theorem: For each t, there is at most one t' such that t steps to t' (t ==> t' is provable). Formally, this is the same as saying that ==> is deterministic.
Proof sketch: We show that if x steps to both y1 and y2 then y1 and y2 are equal, by induction on a derivation of step x y1. There are several cases to consider, depending on the last rule used in this derivation and in the given derivation of step x y2.

Definition deterministic {X: Type} (R: relation X) :=
   x y1 y2 : X, R x y1R x y2y1 = y2.

Theorem step_deterministic:
  deterministic step.
Proof.
  unfold deterministic. intros x y1 y2 Hy1 Hy2.
  generalize dependent y2.
  step_cases (induction Hy1) Case; intros y2 Hy2.
    Case "ST_PlusConstConst". step_cases (inversion Hy2) SCase.
      SCase "ST_PlusConstConst". reflexivity.
      SCase "ST_Plus1". inversion H2.
      SCase "ST_Plus2". inversion H2.
    Case "ST_Plus1". step_cases (inversion Hy2) SCase.
      SCase "ST_PlusConstConst". rewrite <- H0 in Hy1. inversion Hy1.
      SCase "ST_Plus1".
        rewrite <- (IHHy1 t1'0).
        reflexivity. assumption.
      SCase "ST_Plus2". rewrite <- H in Hy1. inversion Hy1.
    Case "ST_Plus2". step_cases (inversion Hy2) SCase.
      SCase "ST_PlusConstConst". rewrite <- H1 in Hy1. inversion Hy1.
      SCase "ST_Plus1". inversion H2.
      SCase "ST_Plus2".
        rewrite <- (IHHy1 t2'0).
        reflexivity. assumption. Qed.

End SimpleArith1.

Values

Let's take a moment to slightly generalize the way we state the definition of single-step reduction.
It is useful to think of the ==> relation as defining an abstract machine:
  • At any moment, the state of the machine is a term.
  • A step of the machine is an atomic unit of computation -- here, a single "add" operation.
  • The halting states of the machine are ones where there is no more computation to be done.
We can then execute a term t as follows:
  • Take t as the starting state of the machine.
  • Repeatedly use the ==> relation to find a sequence of machine states, starting with t, where each state steps to the next.
  • When no more reduction is possible, "read out" the final state of the machine as the result of execution.
Intuitively, it is clear that the final states of the machine are always terms of the form C n for some n. We call such terms values.

Inductive value : tmProp :=
  v_const : n, value (C n).

Having introduced the idea of values, we can use it in the definition of the ==> relation to write ST_Plus2 rule in a slightly more elegant way:

(ST_PlusConstConst) P (C n1) (C n2) ==> C (n1 + n2)
t1 ==> t1'
(ST_Plus1) P t1 t2 ==> P t1' t2
value v1 t2 ==> t2'
(ST_Plus2) P v1 t2 ==> P v1 t2' Again, the variable names here carry important information: by convention, v1 ranges only over values, while t1 and t2 range over arbitrary terms. (Given this convention, the explicit value hypothesis is arguably redundant. We'll keep it for now, to maintain a close correspondence between the informal and Coq versions of the rules, but later on we'll drop it in informal rules, for the sake of brevity.)
Here are the formal rules:

Reserved Notation " t '==>' t' " (at level 40).

Inductive step : tmtmProp :=
  | ST_PlusConstConst : n1 n2,
          P (C n1) (C n2)
      ==> C (n1 + n2)
  | ST_Plus1 : t1 t1' t2,
        t1 ==> t1'
        P t1 t2 ==> P t1' t2
  | ST_Plus2 : v1 t2 t2',
        value v1
        t2 ==> t2'
        P v1 t2 ==> P v1 t2'

  where " t '==>' t' " := (step t t').

Tactic Notation "step_cases" tactic(first) ident(c) :=
  first;
  [ Case_aux c "ST_PlusConstConst"
  | Case_aux c "ST_Plus1" | Case_aux c "ST_Plus2" ].

Exercise: 3 stars (redo_determinism)

As a sanity check on this change, let's re-verify determinism
Proof sketch: We must show that if x steps to both y1 and y2 then y1 and y2 are equal. Consider the final rules used in the derivations of step x y1 and step x y2.
Most of this proof is the same as the one above. But to get maximum benefit from the exercise you should try to write it from scratch and just use the earlier one if you get stuck.

Theorem step_deterministic :
  deterministic step.
Proof.
Admitted.

Strong Progress and Normal Forms

The definition of single-step reduction for our toy language is fairly simple, but for a larger language it would be pretty easy to forget one of the rules and create a situation where some term cannot take a step even though it has not been completely reduced to a value. The following theorem shows that we did not, in fact, make such a mistake here.
Theorem (Strong Progress): If t is a term, then either t is a value, or there exists a term t' such that t ==> t'.
Proof: By induction on t.
  • Suppose t = C n. Then t is a value.
  • Suppose t = P t1 t2, where (by the IH) t1 is either a value or can step to some t1', and where t2 is either a value or can step to some t2'. We must show P t1 t2 is either a value or steps to some t'.

Theorem strong_progress : t,
  value t ( t', t ==> t').
Proof.
  tm_cases (induction t) Case.
    Case "C". left. apply v_const.
    Case "P". right. inversion IHt1.
      SCase "l". inversion IHt2.
        SSCase "l". inversion H. inversion H0.
           (C (n + n0)).
          apply ST_PlusConstConst.
        SSCase "r". inversion H0 as [t' H1].
           (P t1 t').
          apply ST_Plus2. apply H. apply H1.
      SCase "r". inversion H as [t' H0].
           (P t' t2).
          apply ST_Plus1. apply H0. Qed.

This important property is called strong progress, because every term either is a value or can "make progress" by stepping to some other term. (The qualifier "strong" distinguishes it from a more refined version that we'll see in later chapters, called simply "progress.")
The idea of "making progress" can be extended to tell us something interesting about values: in this language values are exactly the terms that cannot make progress in this sense.
To state this observation formally, let's begin by giving a name to terms that cannot make progress. We'll call them normal forms.

Definition normal_form {X:Type} (R:relation X) (t:X) : Prop :=
  ¬ t', R t t'.

This definition actually specifies what it is to be a normal form for an arbitrary relation R over an arbitrary set X, not just for the particular single-step reduction relation over terms that we are interested in at the moment. We'll re-use the same terminology for talking about other relations later in the course.
We can use this terminology to generalize the observation we made in the strong progress theorem: in this language, normal forms and values are actually the same thing.

Lemma value_is_nf : v,
  value vnormal_form step v.
Proof.
  unfold normal_form. intros v H. inversion H.
  intros contra. inversion contra. inversion H1.
Qed.

Lemma nf_is_value : t,
  normal_form step tvalue t.
Proof.   unfold normal_form. intros t H.
  assert (G : value t t', t ==> t').
    SCase "Proof of assertion". apply strong_progress.
  inversion G.
    SCase "l". apply H0.
    SCase "r". apply ex_falso_quodlibet. apply H. assumption. Qed.

Corollary nf_same_as_value : t,
  normal_form step t value t.
Proof.
  split. apply nf_is_value. apply value_is_nf. Qed.

Why is this interesting?
Because value is a syntactic concept -- it is defined by looking at the form of a term -- while normal_form is a semantic one -- it is defined by looking at how the term steps. It is not obvious that these concepts should coincide!
Indeed, we could easily have written the definitions so that they would not coincide...


We might, for example, mistakenly define value so that it includes some terms that are not finished reducing.

Module Temp1.

Inductive value : tmProp :=
| v_const : n, value (C n)
| v_funny : t1 n2,
              value (P t1 (C n2)).

Reserved Notation " t '==>' t' " (at level 40).

Inductive step : tmtmProp :=
  | ST_PlusConstConst : n1 n2,
      P (C n1) (C n2) ==> C (n1 + n2)
  | ST_Plus1 : t1 t1' t2,
      t1 ==> t1'
      P t1 t2 ==> P t1' t2
  | ST_Plus2 : v1 t2 t2',
      value v1
      t2 ==> t2'
      P v1 t2 ==> P v1 t2'

  where " t '==>' t' " := (step t t').

Exercise: 3 stars, advanced (value_not_same_as_normal_form1)

Lemma value_not_same_as_normal_form :
   v, value v ¬ normal_form step v.
Proof.
Admitted.
End Temp1.

Alternatively, we might mistakenly define step so that it permits something designated as a value to reduce further.

Module Temp2.

Inductive value : tmProp :=
| v_const : n, value (C n).

Reserved Notation " t '==>' t' " (at level 40).

Inductive step : tmtmProp :=
  | ST_Funny : n,
      C n ==> P (C n) (C 0)
  | ST_PlusConstConst : n1 n2,
      P (C n1) (C n2) ==> C (n1 + n2)
  | ST_Plus1 : t1 t1' t2,
      t1 ==> t1'
      P t1 t2 ==> P t1' t2
  | ST_Plus2 : v1 t2 t2',
      value v1
      t2 ==> t2'
      P v1 t2 ==> P v1 t2'

  where " t '==>' t' " := (step t t').

Exercise: 2 stars, advanced (value_not_same_as_normal_form2)

Lemma value_not_same_as_normal_form :
   v, value v ¬ normal_form step v.
Proof.
Admitted.

End Temp2.

Finally, we might define value and step so that there is some term that is not a value but that cannot take a step in the step relation. Such terms are said to be stuck. In this case this is caused by a mistake in the semantics, but we will also see situations where, even in a correct language definition, it makes sense to allow some terms to be stuck.

Module Temp3.

Inductive value : tmProp :=
  | v_const : n, value (C n).

Reserved Notation " t '==>' t' " (at level 40).

Inductive step : tmtmProp :=
  | ST_PlusConstConst : n1 n2,
      P (C n1) (C n2) ==> C (n1 + n2)
  | ST_Plus1 : t1 t1' t2,
      t1 ==> t1'
      P t1 t2 ==> P t1' t2

  where " t '==>' t' " := (step t t').

(Note that ST_Plus2 is missing.)

Exercise: 3 stars, advanced (value_not_same_as_normal_form')

Lemma value_not_same_as_normal_form :
   t, ¬ value t normal_form step t.
Proof.
Admitted.

End Temp3.

Additional Exercises


Module Temp4.

Here is another very simple language whose terms, instead of being just plus and numbers, are just the booleans true and false and a conditional expression...

Inductive tm : Type :=
  | ttrue : tm
  | tfalse : tm
  | tif : tmtmtmtm.

Inductive value : tmProp :=
  | v_true : value ttrue
  | v_false : value tfalse.

Reserved Notation " t '==>' t' " (at level 40).

Inductive step : tmtmProp :=
  | ST_IfTrue : t1 t2,
      tif ttrue t1 t2 ==> t1
  | ST_IfFalse : t1 t2,
      tif tfalse t1 t2 ==> t2
  | ST_If : t1 t1' t2 t3,
      t1 ==> t1'
      tif t1 t2 t3 ==> tif t1' t2 t3

  where " t '==>' t' " := (step t t').

Exercise: 1 star (smallstep_bools)

Which of the following propositions are provable? (This is just a thought exercise, but for an extra challenge feel free to prove your answers in Coq.)

Definition bool_step_prop1 :=
  tfalse ==> tfalse.


Definition bool_step_prop2 :=
     tif
       ttrue
       (tif ttrue ttrue ttrue)
       (tif tfalse tfalse tfalse)
  ==>
     ttrue.


Definition bool_step_prop3 :=
     tif
       (tif ttrue ttrue ttrue)
       (tif ttrue ttrue ttrue)
       tfalse
   ==>
     tif
       ttrue
       (tif ttrue ttrue ttrue)
       tfalse.

Exercise: 3 stars, optional (progress_bool)

Just as we proved a progress theorem for plus expressions, we can do so for boolean expressions, as well.

Theorem strong_progress : t,
  value t ( t', t ==> t').
Proof.
Admitted.

Exercise: 2 stars, optional (step_deterministic)

Theorem step_deterministic :
  deterministic step.
Proof.
Admitted.

Module Temp5.

Exercise: 2 stars (smallstep_bool_shortcut)

Suppose we want to add a "short circuit" to the step relation for boolean expressions, so that it can recognize when the then and else branches of a conditional are the same value (either ttrue or tfalse) and reduce the whole conditional to this value in a single step, even if the guard has not yet been reduced to a value. For example, we would like this proposition to be provable: tif (tif ttrue ttrue ttrue) tfalse tfalse ==> tfalse.
Write an extra clause for the step relation that achieves this effect and prove bool_step_prop4.

Reserved Notation " t '==>' t' " (at level 40).

Inductive step : tmtmProp :=
  | ST_IfTrue : t1 t2,
      tif ttrue t1 t2 ==> t1
  | ST_IfFalse : t1 t2,
      tif tfalse t1 t2 ==> t2
  | ST_If : t1 t1' t2 t3,
      t1 ==> t1'
      tif t1 t2 t3 ==> tif t1' t2 t3


  where " t '==>' t' " := (step t t').

Definition bool_step_prop4 :=
         tif
            (tif ttrue ttrue ttrue)
            tfalse
            tfalse
     ==>
         tfalse.

Example bool_step_prop4_holds :
  bool_step_prop4.
Proof.
Admitted.

Exercise: 3 stars, optional (properties_of_altered_step)

It can be shown that the determinism and strong progress theorems for the step relation in the lecture notes also hold for the definition of step given above. After we add the clause ST_ShortCircuit...
  • Is the step relation still deterministic? Write yes or no and briefly (1 sentence) explain your answer.
    Optional: prove your answer correct in Coq.

  • Does a strong progress theorem hold? Write yes or no and briefly (1 sentence) explain your answer.
    Optional: prove your answer correct in Coq.

  • In general, is there any way we could cause strong progress to fail if we took away one or more constructors from the original step relation? Write yes or no and briefly (1 sentence) explain your answer.

End Temp5.
End Temp4.

Multi-Step Reduction

Until now, we've been working with the single-step reduction relation ==>, which formalizes the individual steps of an abstract machine for executing programs.
We can also use this machine to reduce programs to completion -- to find out what final result they yield. This can be formalized as follows:
  • First, we define a multi-step reduction relation ==>*, which relates terms t and t' if t can reach t' by any number of single reduction steps (including zero steps!).
  • Then we define a "result" of a term t as a normal form that t can reach by multi-step reduction.


Since we'll want to reuse the idea of multi-step reduction many times in this and future chapters, let's take a little extra trouble here and define it generically.
Given a relation R, we define a relation multi R, called the multi-step closure of R as follows:

Inductive multi {X:Type} (R: relation X) : relation X :=
  | multi_refl : (x : X), multi R x x
  | multi_step : (x y z : X),
                    R x y
                    multi R y z
                    multi R x z.

The effect of this definition is that multi R relates two elements x and y if either
  • x = y, or else
  • there is some sequence z1, z2, ..., zn such that R x z1 R z1 z2 ... R zn y.
Thus, if R describes a single-step of computation, z1, ... zn is the sequence of intermediate steps of computation between x and y.

Tactic Notation "multi_cases" tactic(first) ident(c) :=
  first;
  [ Case_aux c "multi_refl" | Case_aux c "multi_step" ].

We write ==>* for the multi step relation -- i.e., the relation that relates two terms t and t' if we can get from t to t' using the step relation zero or more times.

Definition multistep := multi step.
Notation " t '==>*' t' " := (multistep t t') (at level 40).

The relation multi R has several crucial properties.
First, it is obviously reflexive (that is, x, multi R x x). In the case of the ==>* (i.e. multi step) relation, the intuition is that a term can execute to itself by taking zero steps of execution.
Second, it contains R -- that is, single-step executions are a particular case of multi-step executions. (It is this fact that justifies the word "closure" in the term "multi-step closure of R.")

Theorem multi_R : (X:Type) (R:relation X) (x y : X),
       R x y → (multi R) x y.
Proof.
  intros X R x y H.
  apply multi_step with y. apply H. apply multi_refl. Qed.

Third, multi R is transitive.

Theorem multi_trans :
   (X:Type) (R: relation X) (x y z : X),
      multi R x y
      multi R y z
      multi R x z.
Proof.
  intros X R x y z G H.
  multi_cases (induction G) Case.
    Case "multi_refl". assumption.
    Case "multi_step".
      apply multi_step with y. assumption.
      apply IHG. assumption. Qed.

That is, if t1==>*t2 and t2==>*t3, then t1==>*t3.

Examples


Lemma test_multistep_1:
      P
        (P (C 0) (C 3))
        (P (C 2) (C 4))
   ==>*
      C ((0 + 3) + (2 + 4)).
Proof.
  apply multi_step with
            (P
                (C (0 + 3))
                (P (C 2) (C 4))).
  apply ST_Plus1. apply ST_PlusConstConst.
  apply multi_step with
            (P
                (C (0 + 3))
                (C (2 + 4))).
  apply ST_Plus2. apply v_const.
  apply ST_PlusConstConst.
  apply multi_R.
  apply ST_PlusConstConst. Qed.

Here's an alternate proof that uses eapply to avoid explicitly constructing all the intermediate terms.

Lemma test_multistep_1':
      P
        (P (C 0) (C 3))
        (P (C 2) (C 4))
  ==>*
      C ((0 + 3) + (2 + 4)).
Proof.
  eapply multi_step. apply ST_Plus1. apply ST_PlusConstConst.
  eapply multi_step. apply ST_Plus2. apply v_const.
  apply ST_PlusConstConst.
  eapply multi_step. apply ST_PlusConstConst.
  apply multi_refl. Qed.

Exercise: 1 star, optional (test_multistep_2)

Lemma test_multistep_2:
  C 3 ==>* C 3.
Proof.
Admitted.

Exercise: 1 star, optional (test_multistep_3)

Lemma test_multistep_3:
      P (C 0) (C 3)
   ==>*
      P (C 0) (C 3).
Proof.
Admitted.

Exercise: 2 stars (test_multistep_4)

Lemma test_multistep_4:
      P
        (C 0)
        (P
          (C 2)
          (P (C 0) (C 3)))
  ==>*
      P
        (C 0)
        (C (2 + (0 + 3))).
Proof.
Admitted.

Normal Forms Again

If t reduces to t' in zero or more steps and t' is a normal form, we say that "t' is a normal form of t."

Definition step_normal_form := normal_form step.

Definition normal_form_of (t t' : tm) :=
  (t ==>* t' step_normal_form t').

We have already seen that, for our language, single-step reduction is deterministic -- i.e., a given term can take a single step in at most one way. It follows from this that, if t can reach a normal form, then this normal form is unique. In other words, we can actually pronounce normal_form t t' as "t' is the normal form of t."

Exercise: 3 stars, optional (normal_forms_unique)

Theorem normal_forms_unique:
  deterministic normal_form_of.
Proof.
  unfold deterministic. unfold normal_form_of. intros x y1 y2 P1 P2.
  inversion P1 as [P11 P12]; clear P1. inversion P2 as [P21 P22]; clear P2.
  generalize dependent y2.
Admitted.
Indeed, something stronger is true for this language (though not for all languages): the reduction of any term t will eventually reach a normal form -- i.e., normal_form_of is a total function. Formally, we say the step relation is normalizing.

Definition normalizing {X:Type} (R:relation X) :=
   t, t',
    (multi R) t t' normal_form R t'.

To prove that step is normalizing, we need a couple of lemmas.
First, we observe that, if t reduces to t' in many steps, then the same sequence of reduction steps within t is also possible when t appears as the left-hand child of a P node, and similarly when t appears as the right-hand child of a P node whose left-hand child is a value.

Lemma multistep_congr_1 : t1 t1' t2,
     t1 ==>* t1'
     P t1 t2 ==>* P t1' t2.
Proof.
  intros t1 t1' t2 H. multi_cases (induction H) Case.
    Case "multi_refl". apply multi_refl.
    Case "multi_step". apply multi_step with (P y t2).
        apply ST_Plus1. apply H.
        apply IHmulti. Qed.

Exercise: 2 stars (multistep_congr_2)

Lemma multistep_congr_2 : t1 t2 t2',
     value t1
     t2 ==>* t2'
     P t1 t2 ==>* P t1 t2'.
Proof.
Admitted.
Theorem: The step function is normalizing -- i.e., for every t there exists some t' such that t steps to t' and t' is a normal form.
Proof sketch: By induction on terms. There are two cases to consider:
  • t = C n for some n. Here t doesn't take a step, and we have t' = t. We can derive the left-hand side by reflexivity and the right-hand side by observing (a) that values are normal forms (by nf_same_as_value) and (b) that t is a value (by v_const).
  • t = P t1 t2 for some t1 and t2. By the IH, t1 and t2 have normal forms t1' and t2'. Recall that normal forms are values (by nf_same_as_value); we know that t1' = C n1 and t2' = C n2, for some n1 and n2. We can combine the ==>* derivations for t1 and t2 to prove that P t1 t2 reduces in many steps to C (n1 + n2).
    It is clear that our choice of t' = C (n1 + n2) is a value, which is in turn a normal form.

Theorem step_normalizing :
  normalizing step.
Proof.
  unfold normalizing.
  tm_cases (induction t) Case.
    Case "C".
       (C n).
      split.
      SCase "l". apply multi_refl.
      SCase "r".
        rewrite nf_same_as_value. apply v_const.
    Case "P".
      inversion IHt1 as [t1' H1]; clear IHt1. inversion IHt2 as [t2' H2]; clear IHt2.
      inversion H1 as [H11 H12]; clear H1. inversion H2 as [H21 H22]; clear H2.
      rewrite nf_same_as_value in H12. rewrite nf_same_as_value in H22.
      inversion H12 as [n1]. inversion H22 as [n2].
      rewrite <- H in H11.
      rewrite <- H0 in H21.
       (C (n1 + n2)).
      split.
        SCase "l".
          apply multi_trans with (P (C n1) t2).
          apply multistep_congr_1. apply H11.
          apply multi_trans with
             (P (C n1) (C n2)).
          apply multistep_congr_2. apply v_const. apply H21.
          apply multi_R. apply ST_PlusConstConst.
        SCase "r".
          rewrite nf_same_as_value. apply v_const. Qed.

Equivalence of Big-Step and Small-Step Reduction

Having defined the operational semantics of our tiny programming language in two different styles, it makes sense to ask whether these definitions actually define the same thing! They do, though it takes a little work to show it. (The details are left as an exercise).

Exercise: 3 stars (eval__multistep)

Theorem eval__multistep : t n,
  t || nt ==>* C n.

The key idea behind the proof comes from the following picture: P t1 t2 ==> (by ST_Plus1) P t1' t2 ==> (by ST_Plus1) P t1'' t2 ==> (by ST_Plus1) ... P (C n1) t2 ==> (by ST_Plus2) P (C n1) t2' ==> (by ST_Plus2) P (C n1) t2'' ==> (by ST_Plus2) ... P (C n1) (C n2) ==> (by ST_PlusConstConst) C (n1 + n2) That is, the multistep reduction of a term of the form P t1 t2 proceeds in three phases:
To formalize this intuition, you'll need to use the congruence lemmas from above (you might want to review them now, so that you'll be able to recognize when they are useful), plus some basic properties of ==>*: that it is reflexive, transitive, and includes ==>.

Proof.
Admitted.

Exercise: 3 stars, advanced (eval__multistep_inf)

Write a detailed informal version of the proof of eval__multistep.
For the other direction, we need one lemma, which establishes a relation between single-step reduction and big-step evaluation.

Exercise: 3 stars (step__eval)

Lemma step__eval : t t' n,
     t ==> t'
     t' || n
     t || n.
Proof.
  intros t t' n Hs. generalize dependent n.
Admitted.
The fact that small-step reduction implies big-step is now straightforward to prove, once it is stated correctly.
The proof proceeds by induction on the multi-step reduction sequence that is buried in the hypothesis normal_form_of t t'. Make sure you understand the statement before you start to work on the proof.

Exercise: 3 stars (multistep__eval)

Theorem multistep__eval : t t',
  normal_form_of t t' n, t' = C n t || n.
Proof.
Admitted.

Additional Exercises

Exercise: 3 stars, optional (interp_tm)

Remember that we also defined big-step evaluation of tms as a function evalF. Prove that it is equivalent to the existing semantics.
Hint: we just proved that eval and multistep are equivalent, so logically it doesn't matter which you choose. One will be easier than the other, though!

Theorem evalF_eval : t n,
  evalF t = n t || n.
Proof.
Admitted.

Exercise: 4 stars (combined_properties)

We've considered the arithmetic and conditional expressions separately. This exercise explores how the two interact.

Module Combined.

Inductive tm : Type :=
  | C : nattm
  | P : tmtmtm
  | ttrue : tm
  | tfalse : tm
  | tif : tmtmtmtm.

Tactic Notation "tm_cases" tactic(first) ident(c) :=
  first;
  [ Case_aux c "C" | Case_aux c "P"
  | Case_aux c "ttrue" | Case_aux c "tfalse" | Case_aux c "tif" ].

Inductive value : tmProp :=
  | v_const : n, value (C n)
  | v_true : value ttrue
  | v_false : value tfalse.

Reserved Notation " t '==>' t' " (at level 40).

Inductive step : tmtmProp :=
  | ST_PlusConstConst : n1 n2,
      P (C n1) (C n2) ==> C (n1 + n2)
  | ST_Plus1 : t1 t1' t2,
      t1 ==> t1'
      P t1 t2 ==> P t1' t2
  | ST_Plus2 : v1 t2 t2',
      value v1
      t2 ==> t2'
      P v1 t2 ==> P v1 t2'
  | ST_IfTrue : t1 t2,
      tif ttrue t1 t2 ==> t1
  | ST_IfFalse : t1 t2,
      tif tfalse t1 t2 ==> t2
  | ST_If : t1 t1' t2 t3,
      t1 ==> t1'
      tif t1 t2 t3 ==> tif t1' t2 t3

  where " t '==>' t' " := (step t t').

Tactic Notation "step_cases" tactic(first) ident(c) :=
  first;
  [ Case_aux c "ST_PlusConstConst"
  | Case_aux c "ST_Plus1" | Case_aux c "ST_Plus2"
  | Case_aux c "ST_IfTrue" | Case_aux c "ST_IfFalse" | Case_aux c "ST_If" ].

Earlier, we separately proved for both plus- and if-expressions...
  • that the step relation was deterministic, and
  • a strong progress lemma, stating that every term is either a value or can take a step.
Prove or disprove these two properties for the combined language.


End Combined.

Small-Step Imp

For a more serious example, here is the small-step version of the Imp operational semantics.
The small-step evaluation relations for arithmetic and boolean expressions are straightforward extensions of the tiny language we've been working up to now. To make them easier to read, we introduce the symbolic notations ==>a and ==>b, respectively, for the arithmetic and boolean step relations.

Inductive aval : aexpProp :=
  av_num : n, aval (ANum n).

We are not actually going to bother to define boolean values, since they aren't needed in the definition of ==>b below (why?), though they might be if our language were a bit larger (why?).

Reserved Notation " t '/' st '==>a' t' " (at level 40, st at level 39).

Inductive astep : stateaexpaexpProp :=
  | AS_Id : st i,
      AId i / st ==>a ANum (st i)
  | AS_Plus : st n1 n2,
      APlus (ANum n1) (ANum n2) / st ==>a ANum (n1 + n2)
  | AS_Plus1 : st a1 a1' a2,
      a1 / st ==>a a1'
      (APlus a1 a2) / st ==>a (APlus a1' a2)
  | AS_Plus2 : st v1 a2 a2',
      aval v1
      a2 / st ==>a a2'
      (APlus v1 a2) / st ==>a (APlus v1 a2')
  | AS_Minus : st n1 n2,
      (AMinus (ANum n1) (ANum n2)) / st ==>a (ANum (minus n1 n2))
  | AS_Minus1 : st a1 a1' a2,
      a1 / st ==>a a1'
      (AMinus a1 a2) / st ==>a (AMinus a1' a2)
  | AS_Minus2 : st v1 a2 a2',
      aval v1
      a2 / st ==>a a2'
      (AMinus v1 a2) / st ==>a (AMinus v1 a2')
  | AS_Mult : st n1 n2,
      (AMult (ANum n1) (ANum n2)) / st ==>a (ANum (mult n1 n2))
  | AS_Mult1 : st a1 a1' a2,
      a1 / st ==>a a1'
      (AMult (a1) (a2)) / st ==>a (AMult (a1') (a2))
  | AS_Mult2 : st v1 a2 a2',
      aval v1
      a2 / st ==>a a2'
      (AMult v1 a2) / st ==>a (AMult v1 a2')

    where " t '/' st '==>a' t' " := (astep st t t').

  Reserved Notation " t '/' st '==>b' t' " (at level 40, st at level 39).

  Inductive bstep : statebexpbexpProp :=
  | BS_Eq : st n1 n2,
      (BEq (ANum n1) (ANum n2)) / st ==>b
      (if (beq_nat n1 n2) then BTrue else BFalse)
  | BS_Eq1 : st a1 a1' a2,
      a1 / st ==>a a1'
      (BEq a1 a2) / st ==>b (BEq a1' a2)
  | BS_Eq2 : st v1 a2 a2',
      aval v1
      a2 / st ==>a a2'
      (BEq v1 a2) / st ==>b (BEq v1 a2')
  | BS_LtEq : st n1 n2,
      (BLe (ANum n1) (ANum n2)) / st ==>b
               (if (ble_nat n1 n2) then BTrue else BFalse)
  | BS_LtEq1 : st a1 a1' a2,
      a1 / st ==>a a1'
      (BLe a1 a2) / st ==>b (BLe a1' a2)
  | BS_LtEq2 : st v1 a2 a2',
      aval v1
      a2 / st ==>a a2'
      (BLe v1 a2) / st ==>b (BLe v1 (a2'))
  | BS_NotTrue : st,
      (BNot BTrue) / st ==>b BFalse
  | BS_NotFalse : st,
      (BNot BFalse) / st ==>b BTrue
  | BS_NotStep : st b1 b1',
      b1 / st ==>b b1'
      (BNot b1) / st ==>b (BNot b1')
  | BS_AndTrueTrue : st,
      (BAnd BTrue BTrue) / st ==>b BTrue
  | BS_AndTrueFalse : st,
      (BAnd BTrue BFalse) / st ==>b BFalse
  | BS_AndFalse : st b2,
      (BAnd BFalse b2) / st ==>b BFalse
  | BS_AndTrueStep : st b2 b2',
      b2 / st ==>b b2'
      (BAnd BTrue b2) / st ==>b (BAnd BTrue b2')
  | BS_AndStep : st b1 b1' b2,
      b1 / st ==>b b1'
      (BAnd b1 b2) / st ==>b (BAnd b1' b2)

  where " t '/' st '==>b' t' " := (bstep st t t').

The semantics of commands is the interesting part. We need two small tricks to make it work:
  • We use SKIP as a "command value" -- i.e., a command that has reached a normal form.
    • An assignment command reduces to SKIP (and an updated state).
    • The sequencing command waits until its left-hand subcommand has reduced to SKIP, then throws it away so that reduction can continue with the right-hand subcommand.
  • We reduce a WHILE command by transforming it into a conditional followed by the same WHILE.
(There are other ways of achieving the effect of the latter trick, but they all share the feature that the original WHILE command needs to be saved somewhere while a single copy of the loop body is being evaluated.)

Reserved Notation " t '/' st '==>' t' '/' st' "
                  (at level 40, st at level 39, t' at level 39).

Inductive cstep : (com × state) → (com × state) → Prop :=
  | CS_AssStep : st i a a',
      a / st ==>a a'
      (i ::= a) / st ==> (i ::= a') / st
  | CS_Ass : st i n,
      (i ::= (ANum n)) / st ==> SKIP / (update st i n)
  | CS_SeqStep : st c1 c1' st' c2,
      c1 / st ==> c1' / st'
      (c1 ;; c2) / st ==> (c1' ;; c2) / st'
  | CS_SeqFinish : st c2,
      (SKIP ;; c2) / st ==> c2 / st
  | CS_IfTrue : st c1 c2,
      IFB BTrue THEN c1 ELSE c2 FI / st ==> c1 / st
  | CS_IfFalse : st c1 c2,
      IFB BFalse THEN c1 ELSE c2 FI / st ==> c2 / st
  | CS_IfStep : st b b' c1 c2,
      b / st ==>b b'
      IFB b THEN c1 ELSE c2 FI / st ==> (IFB b' THEN c1 ELSE c2 FI) / st
  | CS_While : st b c1,
          (WHILE b DO c1 END) / st
      ==> (IFB b THEN (c1;; (WHILE b DO c1 END)) ELSE SKIP FI) / st

  where " t '/' st '==>' t' '/' st' " := (cstep (t,st) (t',st')).

Concurrent Imp

Finally, to show the power of this definitional style, let's enrich Imp with a new form of command that runs two subcommands in parallel and terminates when both have terminated. To reflect the unpredictability of scheduling, the actions of the subcommands may be interleaved in any order, but they share the same memory and can communicate by reading and writing the same variables.

Module CImp.

Inductive com : Type :=
  | CSkip : com
  | CAss : idaexpcom
  | CSeq : comcomcom
  | CIf : bexpcomcomcom
  | CWhile : bexpcomcom
  
  | CPar : comcomcom.

Tactic Notation "com_cases" tactic(first) ident(c) :=
  first;
  [ Case_aux c "SKIP" | Case_aux c "::=" | Case_aux c ";"
  | Case_aux c "IFB" | Case_aux c "WHILE" | Case_aux c "PAR" ].

Notation "'SKIP'" :=
  CSkip.
Notation "x '::=' a" :=
  (CAss x a) (at level 60).
Notation "c1 ;; c2" :=
  (CSeq c1 c2) (at level 80, right associativity).
Notation "'WHILE' b 'DO' c 'END'" :=
  (CWhile b c) (at level 80, right associativity).
Notation "'IFB' b 'THEN' c1 'ELSE' c2 'FI'" :=
  (CIf b c1 c2) (at level 80, right associativity).
Notation "'PAR' c1 'WITH' c2 'END'" :=
  (CPar c1 c2) (at level 80, right associativity).

Inductive cstep : (com × state) → (com × state) → Prop :=
    
  | CS_AssStep : st i a a',
      a / st ==>a a'
      (i ::= a) / st ==> (i ::= a') / st
  | CS_Ass : st i n,
      (i ::= (ANum n)) / st ==> SKIP / (update st i n)
  | CS_SeqStep : st c1 c1' st' c2,
      c1 / st ==> c1' / st'
      (c1 ;; c2) / st ==> (c1' ;; c2) / st'
  | CS_SeqFinish : st c2,
      (SKIP ;; c2) / st ==> c2 / st
  | CS_IfTrue : st c1 c2,
      (IFB BTrue THEN c1 ELSE c2 FI) / st ==> c1 / st
  | CS_IfFalse : st c1 c2,
      (IFB BFalse THEN c1 ELSE c2 FI) / st ==> c2 / st
  | CS_IfStep : st b b' c1 c2,
      b /st ==>b b'
      (IFB b THEN c1 ELSE c2 FI) / st ==> (IFB b' THEN c1 ELSE c2 FI) / st
  | CS_While : st b c1,
      (WHILE b DO c1 END) / st ==>
               (IFB b THEN (c1;; (WHILE b DO c1 END)) ELSE SKIP FI) / st
    
  | CS_Par1 : st c1 c1' c2 st',
      c1 / st ==> c1' / st'
      (PAR c1 WITH c2 END) / st ==> (PAR c1' WITH c2 END) / st'
  | CS_Par2 : st c1 c2 c2' st',
      c2 / st ==> c2' / st'
      (PAR c1 WITH c2 END) / st ==> (PAR c1 WITH c2' END) / st'
  | CS_ParDone : st,
      (PAR SKIP WITH SKIP END) / st ==> SKIP / st
  where " t '/' st '==>' t' '/' st' " := (cstep (t,st) (t',st')).

Definition cmultistep := multi cstep.

Notation " t '/' st '==>*' t' '/' st' " :=
   (multi cstep (t,st) (t',st'))
   (at level 40, st at level 39, t' at level 39).

Among the many interesting properties of this language is the fact that the following program can terminate with the variable X set to any value...

Definition par_loop : com :=
  PAR
    Y ::= ANum 1
  WITH
    WHILE BEq (AId Y) (ANum 0) DO
      X ::= APlus (AId X) (ANum 1)
    END
  END.

In particular, it can terminate with X set to 0:

Example par_loop_example_0:
   st',
       par_loop / empty_state ==>* SKIP / st'
     st' X = 0.
Proof.
  eapply ex_intro. split.
  unfold par_loop.
  eapply multi_step. apply CS_Par1.
    apply CS_Ass.
  eapply multi_step. apply CS_Par2. apply CS_While.
  eapply multi_step. apply CS_Par2. apply CS_IfStep.
    apply BS_Eq1. apply AS_Id.
  eapply multi_step. apply CS_Par2. apply CS_IfStep.
    apply BS_Eq. simpl.
  eapply multi_step. apply CS_Par2. apply CS_IfFalse.
  eapply multi_step. apply CS_ParDone.
  eapply multi_refl.
  reflexivity. Qed.

It can also terminate with X set to 2:

Example par_loop_example_2:
   st',
       par_loop / empty_state ==>* SKIP / st'
     st' X = 2.
Proof.
  eapply ex_intro. split.
  eapply multi_step. apply CS_Par2. apply CS_While.
  eapply multi_step. apply CS_Par2. apply CS_IfStep.
    apply BS_Eq1. apply AS_Id.
  eapply multi_step. apply CS_Par2. apply CS_IfStep.
    apply BS_Eq. simpl.
  eapply multi_step. apply CS_Par2. apply CS_IfTrue.
  eapply multi_step. apply CS_Par2. apply CS_SeqStep.
    apply CS_AssStep. apply AS_Plus1. apply AS_Id.
  eapply multi_step. apply CS_Par2. apply CS_SeqStep.
    apply CS_AssStep. apply AS_Plus.
  eapply multi_step. apply CS_Par2. apply CS_SeqStep.
    apply CS_Ass.
  eapply multi_step. apply CS_Par2. apply CS_SeqFinish.

  eapply multi_step. apply CS_Par2. apply CS_While.
  eapply multi_step. apply CS_Par2. apply CS_IfStep.
    apply BS_Eq1. apply AS_Id.
  eapply multi_step. apply CS_Par2. apply CS_IfStep.
    apply BS_Eq. simpl.
  eapply multi_step. apply CS_Par2. apply CS_IfTrue.
  eapply multi_step. apply CS_Par2. apply CS_SeqStep.
    apply CS_AssStep. apply AS_Plus1. apply AS_Id.
  eapply multi_step. apply CS_Par2. apply CS_SeqStep.
    apply CS_AssStep. apply AS_Plus.
  eapply multi_step. apply CS_Par2. apply CS_SeqStep.
    apply CS_Ass.

  eapply multi_step. apply CS_Par1. apply CS_Ass.
  eapply multi_step. apply CS_Par2. apply CS_SeqFinish.
  eapply multi_step. apply CS_Par2. apply CS_While.
  eapply multi_step. apply CS_Par2. apply CS_IfStep.
    apply BS_Eq1. apply AS_Id.
  eapply multi_step. apply CS_Par2. apply CS_IfStep.
    apply BS_Eq. simpl.
  eapply multi_step. apply CS_Par2. apply CS_IfFalse.
  eapply multi_step. apply CS_ParDone.
  eapply multi_refl.
  reflexivity. Qed.

More generally...

Exercise: 3 stars, optional

Lemma par_body_n__Sn : n st,
  st X = n st Y = 0 →
  par_loop / st ==>* par_loop / (update st X (S n)).
Proof.
Admitted.

Exercise: 3 stars, optional

Lemma par_body_n : n st,
  st X = 0 st Y = 0 →
   st',
    par_loop / st ==>* par_loop / st' st' X = n st' Y = 0.
Proof.
Admitted.
... the above loop can exit with X having any value whatsoever.

Theorem par_loop_any_X:
   n, st',
    par_loop / empty_state ==>* SKIP / st'
     st' X = n.
Proof.
  intros n.
  destruct (par_body_n n empty_state).
    split; unfold update; reflexivity.

  rename x into st.
  inversion H as [H' [HX HY]]; clear H.
   (update st Y 1). split.
  eapply multi_trans with (par_loop,st). apply H'.
  eapply multi_step. apply CS_Par1. apply CS_Ass.
  eapply multi_step. apply CS_Par2. apply CS_While.
  eapply multi_step. apply CS_Par2. apply CS_IfStep.
    apply BS_Eq1. apply AS_Id. rewrite update_eq.
  eapply multi_step. apply CS_Par2. apply CS_IfStep.
    apply BS_Eq. simpl.
  eapply multi_step. apply CS_Par2. apply CS_IfFalse.
  eapply multi_step. apply CS_ParDone.
  apply multi_refl.

  rewrite update_neq. assumption. intro X; inversion X.
Qed.

End CImp.

A Small-Step Stack Machine

Last example: a small-step semantics for the stack machine example from Imp.v.

Definition stack := list nat.
Definition prog := list sinstr.

Inductive stack_step : stateprog × stackprog × stackProp :=
  | SS_Push : st stk n p',
    stack_step st (SPush n :: p', stk) (p', n :: stk)
  | SS_Load : st stk i p',
    stack_step st (SLoad i :: p', stk) (p', st i :: stk)
  | SS_Plus : st stk n m p',
    stack_step st (SPlus :: p', n::m::stk) (p', (m+n)::stk)
  | SS_Minus : st stk n m p',
    stack_step st (SMinus :: p', n::m::stk) (p', (m-n)::stk)
  | SS_Mult : st stk n m p',
    stack_step st (SMult :: p', n::m::stk) (p', (m×n)::stk).

Theorem stack_step_deterministic : st,
  deterministic (stack_step st).
Proof.
  unfold deterministic. intros st x y1 y2 H1 H2.
  induction H1; inversion H2; reflexivity.
Qed.

Definition stack_multistep st := multi (stack_step st).

Exercise: 3 stars, advanced (compiler_is_correct)

Remember the definition of compile for aexp given in the Imp chapter. We want now to prove compile correct with respect to the stack machine.
State what it means for the compiler to be correct according to the stack machine small step semantics and then prove it.

Definition compiler_is_correct_statement : Prop :=
admit.

Theorem compiler_is_correct : compiler_is_correct_statement.
Proof.
Admitted.