• Keine Ergebnisse gefunden

Consider the Petri net below, describing Lamport’s 1-bit mutual exclusion algorithm.

N/A
N/A
Protected

Academic year: 2021

Aktie "Consider the Petri net below, describing Lamport’s 1-bit mutual exclusion algorithm."

Copied!
2
0
0

Wird geladen.... (Jetzt Volltext ansehen)

Volltext

(1)

Concurrency Theory (WS 2011/12) Out: Tue, Nov 15 Due: Mon, Nov 21

Exercise Sheet 5

Jun.-Prof. Roland Meyer, Georgel C˘alin Technische Universit¨at Kaiserslautern

Problem 1: Lamport’s Mutual Exclusion Algorithm

Consider the Petri net below, describing Lamport’s 1-bit mutual exclusion algorithm.

cs

1

idle

1

req

1

nid

1

id

1

af ter you

2

await

2

id

2

idle

2

cs

2

req

2

(a) set up the colinear property one would want the mutex to satisfy and determine the connectivity and trap matrices of the given Petri net;

(b) prove that the basic verification system is feasible;

(c) prove that the enhanced verification system is infeasible.

How do you interpret the fact that bvs is feasable and evs infeasible?

The exercise is a bit of work but demonstrates the power of the analysis technique. The example is taken from J. Esparza, S. Melzer: Verification of Safety Properties Using Integer Program- ming: Beyond the State Equation. Formal Methods in System Design 16(2): 159-189 (2000).

Problem 2: Conflict vs. Causality vs. Concurrency

Prove that for occurrence net (B, E, G) and distinct x, y ∈ B ∪E, precisely one of the following holds:

• x and y are causally related (i.e. x < y or x > y)

• x and y are in conflict (i.e. x ] y)

• x and y are concurrent (i.e. x co y)

(2)

Problem 3: Configurations and Firing Sequences

Let C = {e

1

, . . . , e

n

} be a configuration with normal event ordering: i < j if e

i

< e

j

. Prove that e

1

. . . e

n

is a possible firing sequence in the unfolding, producing the marking ({e

}

∪ C

)\

C.

Problem 4: Unfoldings, Configurations, and Cuts

Consider the Petri net depicted below:

(a) unfold the Petri net and outline one of its prefixes;

(b) describe/list all the configurations and cuts of your unfolding.

p

1

p

2

t

1

t

2

p

3

t

3

p

4

p

5

t

4

p

6

p

7

Why is your unfolding finite?

Referenzen

ÄHNLICHE DOKUMENTE

(a) Set up the co-linear property one would want the mutex to satisfy and determine the connectivity and trap matrices of the given Petri net. (b) Prove that the basic

A labelled Petri net is a tuple N = (S, T, W, M 0 , X, λ, F ) where S, T, W and M 0 are the finite set of places, transitions, the weight function and the initial marking

(a) Let load, store and inc be atomic operations for loading a variable’s value into cache, storing a variable’s cache value into memory and incrementing a variable’s cache value. If

In both cases provide the initial marking of the additional places you might have added and argument why your Petri nets do what

Add arcs to the Petri net N below so that its family of generating traps contains exponentially (in N ’s size)

Provide the set of possible extensions and cut-offs at each iteration.. What do

Firing t2 initializes all source links as well as state places, the inner activity, the event handlers, and for each scope embedded in the process its respective place !push

Deduce from Remark 35.12 of the Lecture Notes that column orthogonality relations for the Brauer characters take the form Π tr Φ =