Newtonian Program Analysis: Solving Sharir and Pnueli s Equations

Transcription

1 Newtonian Program Analysis: Solving Sharir and Pnueli s Equations Javier Esparza Technische Universität München Joint work with Stefan Kiefer and Michael Luttenberger

2

3

4 From programs to equations: Intraprocedural x > 0 x = 0 x x + 1 x < 10 x 10

5 From programs to equations: Intraprocedural c Y a d X b One-step relations a,..., e (IN IN) c = {(x, x + 1) x 0} Big-step relations X, Y, Z IN IN Z e X Y = a Y + b = c Z Z = d Y + e

6 From programs to equations: Intraprocedural c Y a d X b One-step relations a,..., e (IN IN) c = {(x, x + 1) x 0} Big-step relation X IN IN Z e X Y = a Y + b = c Z Z = d Y + e

7 From programs to equations: Intraprocedural c Y a d X b One-step relations a,..., e (IN IN) c = {(x, x + 1) x 0} Big-step relations X, Y, Z IN IN Z e X Y = a Y + b = c Z Z = d Y + e

8 From programs to equations: Intraprocedural c Y a d X b One-step relations a,..., e (IN IN) c = {(x, x + 1) x 0} Big-step relations X, Y, Z IN IN Z e X Y = a Y + b = c Z Z = d Y + e

9 From programs to equations: Intraprocedural Program system X = f (X) of linear fixed-point equations Least solution non-computable in general Program analysis: domain 2 IN abstract domain D transformer f abstract transformer f # Sufficient condition for existence of least solution: (D, +, ) is a (ω-continuous) semiring

10 From programs to equations: Intraprocedural Program system X = f (X) of linear fixed-point equations Least solution non-computable in general Program analysis: domain 2 IN abstract domain D transformer f abstract transformer f # Sufficient condition for existence of least solution: (D, +, ) is a (ω-continuous) semiring

11 From programs to equations: Intraprocedural Program system X = f (X) of linear fixed-point equations Least solution non-computable in general Program analysis: domain 2 IN abstract domain D transformer f abstract transformer f # Sufficient condition for existence of least solution: (D, +, ) is a (ω-continuous) semiring

12 From programs to equations: Intraprocedural Program system X = f (X) of linear fixed-point equations Least solution non-computable in general Program analysis: domain 2 IN abstract domain D transformer f abstract transformer f # Sufficient condition for existence of least solution: (D, +, ) is a (ω-continuous) semiring

13 Quantitative program analysis: Expected time X 0.7 Y X = 0.7 Y Y = Z + 1 Z Z = 0.6 Y

14 From programs to equations: Interprocedural P Q x > 0 x = 0 x < 3 x 3 call Q x < 10 call Q call P x 10 x x 2

15 From programs to equations: Interprocedural P Q P 0 a e Q 0 f P 0 = a P 1 + b P 1 =?? P 2 P 1 Q 1 Q 2 P 2 = c P 1 + d call Q c b call Q call P Q 0 = e Q 1 + f Q 2 P 2 d g Q 3 Q 1 =?? Q 3 Q 2 =?? Q 3 Q 3 = g

16 From programs to equations: Interprocedural P Q P 0 = a P 1 + b P 0 Q 0 P 1 =?? P 2 a e f P 2 = c P 1 + d P 1 Q 1 Q 2 Q 0 = e Q 1 + f Q 2 call Q c b call Q call P Q 1 =?? Q 3 P 2 d g Q 3 Q 2 =?? Q 3 Q 3 = g

17 Sharir and Pnueli s functional approach P Q P 0 a e Q 0 f P 0 = a P 1 + b Q 0 P 1 = Q 0 P 2 P 2 = c P 1 + d P 1 Q 1 Q 2 Q 0 = e Q 1 + f Q 2 call Q c b call Q call P Q 1 =?? Q 3 P 2 d g Q 3 Q 2 =?? Q 3 Q 3 = g

18 Sharir and Pnueli s functional approach P Q P 0 a e Q 0 f P 0 = a P 1 + b Q 0 P 1 = Q 0 P 2 P 2 = c P 1 + d P 1 Q 1 Q 2 Q 0 = e Q 1 + f Q 2 call Q c b call Q call P Q 0 Q 1 =?? Q 3 P 2 d g Q 3 Q 2 =?? Q 3 Q 3 = g

19 Sharir and Pnueli s functional approach P Q P 0 a e Q 0 f P 0 = a P 1 + b Q 0 P 1 = Q 0 P 2 P 2 = c P 1 + d P 1 Q 1 Q 2 Q 0 = e Q 1 + f Q 2 call Q c b call Q call P Q 0 Q 1 =?? Q 3 P 2 d g Q 3 Q 2 =?? Q 3 Q 3 = g P 0

20 Sharir and Pnueli s interprocedural equations Program system X = f (X) of polynomial, non-linear fixed-point equations Least solution non-computable in general Program analysis: domain 2 IN abstract domain D transformer f abstract transformer f # Sufficient condition for existence of least solution: (D, +, ) is a (ω-continuous) semiring

21 Sharir and Pnueli s interprocedural equations Program system X = f (X) of polynomial, non-linear fixed-point equations Least solution non-computable in general Program analysis: domain 2 IN abstract domain D transformer f abstract transformer f # Sufficient condition for existence of least solution: (D, +, ) is a (ω-continuous) semiring

22 Solving the equations: Kleene iteration Theorem [Kleene]: The least solution µf of X = f (X) is the supremum of {k i } i 0, where k 0 = f (0) k i+1 = f (k i ) Basic algorithm: compute k 0, k 1, k 2,... until either k i = k i+1, which implies k i = µf, or the approximation is considered adequate.

23 Kleene iteration is slow Set domains: Kleene iteration never terminates for X = f (X) if least solution µf is an infinite set. X = a X + b µf = a b Kleene approximants are finite sets: k i = (ɛ + a a i )b Probabilistic interpretation: convergence can be very slow for polynomial equations [EY STACS05]. X = 1 2 X µf = 1 = Logarithmic convergence : k iterations to get log k bits of accuracy. k n 1 1 n + 1 k 2000 =

24 Kleene Iteration for X = f (X) (univariate case) µf f (X)

25 Kleene Iteration for X = f (X) (univariate case) µf f (X)

26 Kleene Iteration for X = f (X) (univariate case) µf f (X)

27 Kleene Iteration for X = f (X) (univariate case) µf f (X)

28 Kleene Iteration for X = f (X) (univariate case) µf f (X)

29 Kleene Iteration for X = f (X) (univariate case) µf f (X)

30 Kleene Iteration for X = f (X) (univariate case) µf f (X)

31 Kleene Iteration for X = f (X) (univariate case) µf f (X)

32 Newton s Method for X = f (X) (univariate case) µf f (X)

33 Newton s Method for X = f (X) (univariate case) µf f (X)

34 Newton s Method for X = f (X) (univariate case) µf f (X)

35 Newton s Method for X = f (X) (univariate case) µf f (X)

36 Newton s Method for X = f (X) (univariate case) µf f (X)

37 Newton s Method for X = f (X) (univariate case) µf f (X)

38 Kleene vs. Newton Program analysis: Kleene iteration is applicable to arbitrary ω-continuous semirings... but converges slowly. Numerical mathematics: Newton s Method converges fast... but can only be applied to the real field Can Newton s method be generalized to arbitrary ω-continuous semirings?

39 Kleene vs. Newton Program analysis: Kleene iteration is applicable to arbitrary ω-continuous semirings... but converges slowly. Numerical mathematics: Newton s Method converges fast... but can only be applied to the real field Can Newton s method be generalized to arbitrary ω-continuous semirings?

40 Mathematical formulation of Newton s Method Elementary analysis yields for the i-th Newton iterant ν i : ν 0 = 0 ν i+1 = ν i + i where i least solution of X = Df νi (X) + f (ν i ) ν i Df νi (X) differential of f (X) at the point ν i

41 Generalizing Newton s method Key point: generalize X = Df ν (X) + f (ν) ν to arbitrary ω-continuous semirings In an arbitrary ω-continuous semiring neither the differential Df ν (X), nor the difference f (ν) ν are defined.

42 Overcoming the obstacles (1) Use the algebraic definition of differential (recall that we only have polynomial functions!) 0 if f (X) = c X if f (X) = X Df (X) = Dg(X) + Dh(X) if f (X) = g(x) + h(x) Dg(X) h(x) + g(x) Dh(X) if f (X) = g(x) h(x) (2) Replace f (ν i ) ν i by any δ i such that f (ν i ) = ν i + δ i Define i as the least solution of X = Df νi (X) + δ i

43 Idempotent and commutative semirings Theorem [Hopkins-Kozen LICS 99]: The least fixed point of a system X = f (X) of n equations over an idempotent and commutative ω-continuous semiring is reached by the sequence ν 0 = f (0) ν i+1 = J(ν i ) f (ν i ) after at most O(3 n ) iterations. Theorem: This is exactly Newton s sequence. Moreover, the fixed point is reached after at most n iterations.

44 Idempotent and commutative semirings Theorem [Hopkins-Kozen LICS 99]: The least fixed point of a system X = f (X) of n equations over an idempotent and commutative ω-continuous semiring is reached by the sequence ν 0 = f (0) ν i+1 = J(ν i ) f (ν i ) after at most O(3 n ) iterations. Theorem: This is exactly Newton s sequence. Moreover, the fixed point is reached after at most n iterations.

45 Two (quick) examples of application

46 May-Alias Analysis listify(): transforms a binary tree of pointers into a list of pointers by reading the tree in preorder. void listifyl() { void listify() { T.move left(); L.push back(t get data()); listify(); if( T.is leaf() == false ) { T.move up(); listifyl(); listifyr(); } } }

47 May-Alias Analysis

48 May-Alias Analysis

49 May-Alias Analysis Which data access paths of tree and list may point to the same element? Access path of the tree: word w 1 {left, right} Access path of the list: word w 2 {next} May-alias information: set of pairs (w 1, w 2 ) such that w 1, w 2 may point to the same cell Commutative abstraction: (left right left, next next) (3, 1, 2)

50 May-Alias Analysis Kleene iteration does not terminate. Newton s method terminates after one iteration and provides the following information: Data access paths with 0 right s and l left s may only be aliased to the l-th element of the list. Data access paths with r right s and l left s may only be aliased to the (2r + l)-th element of the list, or to larger elements of the same parity.

51 Lazy evaluation of And-Or trees Nodes are only constructed and evaluated (to 0 or 1) if needed. (e.g., if left subtree of And-node evaluates to 1, right subtree is not constructed) function And(node) if node.leaf() then return node.value() else v := Or(node.left) if v = 0 then return 0 else return Or(node.right) function Or(node) if node.leaf() then return node.value() else v := And(node.left) if v = 1 then return 1 else return And(node.right)

52 Assume the probabilities that node.leaf() returns true and node.value() returns 1 are both 1/2. We perform an analysis to compute the average runtime.

53 Kleene vs. Newton Neither Kleene nor Newton terminate, but Newton converges faster: i k (i) And 0 ν (i) And 0 k (i) And 1 ν (i) And

54 Messages of this talk The theory of solving program analysis equations is not be as well understood as we thought. We can learn from numerical mathematics and contribute to it. Go for unified theory of qualitative and quantitative program analysis.

55 Commenting the works...

56 ... of the giants

57 And yet... The glosses of Saint Emilianus, written around 1000 AC. Oldest writing in (a form of) Spanish, one of the most spoken languages in the world.

58 Shimon Even, Amir Pnueli,