1 / 17

Hoare ’ s method for proving correctness of programs

Hoare ’ s method for proving correctness of programs. Hoare ’ s logic. Hoare suggested in 1969 an inference system with which we can prove properties of programs. A unique form of proof rules, called Hoare triples: P {Q} R where P is a precondition Q is a program or a program statement

ishmael
Télécharger la présentation

Hoare ’ s method for proving correctness of programs

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript


  1. Hoare’s method for proving correctness of programs

  2. Hoare’s logic • Hoare suggested in 1969 an inference system with which we can prove properties of programs. • A unique form of proof rules, called Hoare triples:P {Q} Rwhere • P is a precondition • Q is a program or a program statement • R postcondition.

  3. The simple language we consider • Hoare’s logic includes rules for inference on program constructs: • Assignments y = t • Composition S1; S2 • If-then-else if e then S1 else S2 fi • While while e do S od • ...

  4. The simple language we consider • A proof typically needs additional inference rules for the elementary operations (+,-,*,mod...) • e.g., axioms for arithmetic such as • x + 0 = 0 • x £ y = y £ x • ...

  5. Assignments • D0 Axiom of assignments P {x = f} R where • P is derived from R by replacing all occurrences of x with f. • Example: y + 1 < 10 {y = y+1} y < 10

  6. Consequence • D1 Inference rules for consequence P {Q} R R → S (make a postcondition weaker) P {Q} S P {Q} R S → P (make a precondition stronger) S {Q} R • Example y + 1 < 10 {y = y+1} y < 10 y < 10 → x = 1 y + 1 < 10 {y = y+1} x = 1

  7. Composition • D2 Inference rule for composition P {Q1} R1 R1 {Q2} R P {Q1; Q2} R • Example: x=1 {y = x} y=1 y=1 {z=y} z=1 x=1{y=x; z=y} z=1

  8. Iteration • Consider the statement: while B do S • D3 Inference rule for iteration P B {S} P P {while B do S}:B Æ P • Example x < 10 x < 5 {x = x+1} x < 10x<10 {while (x < 5) x = x + 1} x ¸ 5 Æ x < 10

  9. if-then-else • D4 if-then-else inference rule P Æ e {Q1} R P Æ ¬e {Q2} R P {if e then Q1 else Q2 fi} R • Example z = 0 Æ x < 0 {y := -x + z} y ¸ 0 z = 0 Æ x ¸ 0 {y := x + z} y ¸ 0 z = 0 {if (x<0) y := -x + z else y := x + z} y ¸ 0

  10. Example • Consider the following program for finding the quotient and remainder of x / y. r := x; q := 0; while (y <= r) { r := r – y; q := 1 + q;} • We will try to prove that in the end: x = r + y£ q and y > r. x = 16, y = 6

  11. Example • Strategy: find a loop invariant. r := x; q := 0; while (y <= r) { r := r – y; q := 1 + q;} x = x + y £ 0 x = r + y £ 0 x = r + y £ q

  12. Proof • So we first need to prove Lemma 1: x = x + y £ 0 • For this we will use the following axioms from arithmetic: • A0 x = x • A1 x + 0 = x • A2 x £ 0 = 0 • Proof • x = x A0 • y £ 0 = 0 A2 • x = x + y £ 0 1,2,A1

  13. Proof • true → x = x + y £ 0 Lemma1 • x = x + y £ 0{r := x}x = r + y £ 0D0 • x = r + y £ 0{q := 0}x = r + y £ qD0 • true{r := x}x = r + y £ 0D1, 1,2 • true{r := x; q :=0}x = r + y £ qD2, 4,3

  14. ! ! Proof Line Formal proof Justification (Continued on next page ...)

  15. Proof (cont’d) Line Formal proof Justification

  16. Notes on Hoare-style rules • Proves partial correctness • i.e., meaningless when the program does not terminate. • If in addition we prove termination, then we have total correctness. • Q: suppose that y < 0 in our program. Does it change the proof ?

  17. Notes on Hoare-style rules • Hoare logic is sound in the sense that everything that can be proved is correct • There exists extensions to all popular constructs: • recursion • arrays, pointers,... • concurrency...

More Related