Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Investigating Deadlocks

Download as pdf or txt
Download as pdf or txt
You are on page 1of 4
At a glance
Powered by AI
The key takeaways are how to construct a deadlock condition, recover from deadlocks, and prevent deadlocks using different techniques.

A deadlock condition can be constructed by having four processes allocate different resources and later request another resource, creating a cycle in the resource allocation graph.

Two methods of recovering from a deadlock condition are releasing a held resource or removing a waiting process.

Investigating Deadlocks

Introduction
Objectives
At the end of this lab you should be able to:

Construct a resource allocation graph for a deadlock condition and verify


using the simulator.

Use two methods of resolving a deadlock condition.

Use two methods of preventing a deadlock condition.

Explain and use the total ordering method to prevent a deadlock.

Basic Theory
Process deadlocks which are created when processes require
access to more than one resource (not-sharable) are frequent
occurrences in a modern operating system (OS). The OS can
employ various techniques for avoiding, preventing or resolving
process deadlocks which will allow the system to run as efficiently
as possible.

Lab Exercises - Investigate and Explore

1. Four processes are running. They are called P1 to P4.There are also four
resources available (only one instance of each). They are named R0 to R3.
At some point of their existence each process allocates a different resource
for use and holds it for itself forever. Later each of the processes request
another one of the four resources. Draw the resource allocation graph for a
four process deadlock condition. Do not continue until you do this and get it
verified by the tutor.

2. In the compiler window, enter the following source code in the compiler
source editor area (under PROGRAM SOURCE frame title).

program DeadlockPN
resource(X, allocate)
wait(3)
resource(Y, allocate)
for n = 1 to 20
next
end
The above code creates a program which attempts to allocate two
resources for itself. After the first allocation it waits for 3 seconds and tries
to allocate another resource. Finally it counts from 1 to 20 in a loop and
then terminates. Other than that it does nothing sensible or useful!
Now follow the instructions below as faithfully as you can:
a. Copy the above code and paste it in three more edit windows so that
you have a total of four pieces of source code.
b. In each case change N in the program name to 1 to 4, e.g.
DeadlockP1, DeadlockP2, etc.
c. Look at your graph you constructed in (1) above and using that
information fill in the values for each of the Xs and Ys in the four
pieces of source code.
d. Compile each one of the four source code.
e. Load in memory the four pieces of code generated.
f. Now switch to the OS simulator.
g. Create a single instance of each of the programs. You can do this by
double-clicking on each of the program names in the PROGRAM LIST
frame under the Program Name column.
h. In the SCHEDULER frame select Round Robin (RR) scheduling
policy in the Policies tab.
i. In OS Control tab, push the speed slider up to the fastest speed.
j. Select the Views tab and click on the VIEW RESOURCES button.
k. Select Stay on top check box in the displayed window.
l. Back in the OS Control tab use the START button to start the OS
scheduler and observe the changing process states for few seconds.
m. Have you got a deadlock condition same as you constructed in (1)
above? If you havent then check and if necessary re-do above. Do not
proceed to (n) or (3) below until you get a deadlock condition.
n. If you have a deadlock condition then click on the SHOW
DEADLOCKED PROCESSES button in the System Resources

window. Does the highlighted resource allocation graph look like


yours?
3. Now that you created a deadlock condition let us try two methods of getting
out of this condition:
a. In the System Resources window, there should be four resource shapes
that are in red colour indicating they are both allocated to one process
and requested by another.
b. Select one of these resources and click on the Release button next to it.
c. Observe what is happening to the processes in the OS Simulator window.
d. Is the deadlock situation resolved? Explain briefly why this helped resolve
the deadlock.
e. Re-create the same deadlock condition (steps in 2 above should help).
f. Once the deadlock condition is obtained again do the following: In the OS
Simulator window, select a process in the waiting queue in the WAITING
PROCESSES frame.
g. Click on the REMOVE button and observe the processes.
h. Has this managed to resolve the deadlock? Explain briefly why this helped
resolve the deadlock.
This part of the exercises was about two methods of recovering from a
deadlock condition after it happens.
4. We now look at two methods of preventing a deadlock condition before it
happens.
a. In the System Resources window select the Disallow hold and wait
check box in the Prevent frame.
b. Try to re-create the same deadlock condition as before. Have you been
successful? What happened? Click on the SHOW DEADLOCKED
PROCESSES button and observe the displayed information in the text
window for potential clues.
c. Next, uncheck the Disallow hold and wait check box and check the
Disallow circular wait check box.
d. Try to re-create the same deadlock condition as before. Have you been
successful? What happened? Click on the SHOW DEADLOCKED
PROCESSES button and observe the displayed information in the text
window for potential clues.
5. We are now going to try a third method of preventing deadlocking before it
happens. It is called total ordering method. Here the resources are
allocated in increasing resource id numbers only. So, for example, resource
R3 must be allocated after resources R0 to R2 and resource R1 cannot be

allocated after resource R2 is allocated. Looking at your resource allocation


graph can you see how this ordering can prevent a deadlock? Comment.
a. In the System Resources window select the Use total ordering check
box in the Prevent frame. The other options should be unchecked.
b. Try to re-create the same deadlock condition as before. Have you been
successful? What happened? Click on the SHOW DEADLOCKED
PROCESSES button and observe the displayed information in the text
window for potential clues. What happened? Comment.

You might also like