Chapter 7: Deadlocks

Chapter 7: Deadlocks

 The Deadlock Problem

 System Model
 Deadlock Characterization
 Methods for Handling Deadlocks
 Deadlock Prevention
 Deadlock Avoidance
 Deadlock Detection
 Recovery from Deadlock

Operating System Concepts - 7th Edition, Feb 14, 2005 7.2 Silberschatz, Galvin and Gagne ©2005
 Chapter Objectives

 To develop a description of deadlocks, which

prevent sets of concurrent processes from
completing their tasks
 To present a number of different methods for
preventing or avoiding deadlocks in a computer
system.

Operating System Concepts - 7th Edition, Feb 14, 2005 7.3 Silberschatz, Galvin and Gagne ©2005
 The Deadlock Problem

 In multiprogramming environment, several processes

may compete for a finite (limited) number of resources. If
the resources are not available at that time, then they
wait, they continue to wait until the resources are
available.
 A set of blocked processes each holding a resource and
waiting to acquire a resource held by another process in
the set.
 Example
 System has 2 disk drives.
 P1 and P2 each hold one disk drive and each needs
another one.

Operating System Concepts - 7th Edition, Feb 14, 2005 7.4 Silberschatz, Galvin and Gagne ©2005
 Bridge Crossing Example

 Traffic only in one direction.

 Each section of a bridge can be viewed as a resource.
 If a deadlock occurs, it can be resolved if one car backs
up (preempt resources and rollback).
 Several cars may have to be backed up if a deadlock
occurs.
 Starvation is possible.

Operating System Concepts - 7th Edition, Feb 14, 2005 7.5 Silberschatz, Galvin and Gagne ©2005
 System Model
 A system consists of a finite number of resources to be distributed among
a number of competing processes.
 Resource types R1, R2, . . ., Rm
CPU cycles, memory space, I/O devices
 Each resource type Ri has Wi instances (ex: if we have two CPU’s in the
computer then we have two instances from the CPU).
 Each process utilizes a resource as follows (mode of operation):
 Request : if the request cannot be granted immediately (for example
the resource is being used by another process), then the requesting
process must wait until it can acquire the resource.
 Use:
 Release
 Both release and request of resources are system calls.
 A system table used to record weather each resource is free or allocated,
and to which process the resource is allocated.

Operating System Concepts - 7th Edition, Feb 14, 2005 7.6 Silberschatz, Galvin and Gagne ©2005
 Deadlock Characterization

Deadlock can arise if four conditions hold simultaneously (all of them together).
 Mutual exclusion: only one process at a time can use a resource (the
resource have one instance or no sharing is allowed for that resource
(one process at a time can use that resource)).
 Hold and wait: a process holding at least one resource is waiting to
acquire additional resources held by other processes.
 No preemption: a resource can be released only voluntarily by the
process holding it, after that process has completed its task.
 Circular wait: there exists a set {P , P , …, P } of waiting processes
0 1 0
such that P0 is waiting for a resource that is held by P1, P1 is waiting for
a resource that is held by P2, …, Pn–1 is waiting for a resource that is
held by Pn, and P0 is waiting for a resource that is held by P0.

Operating System Concepts - 7th Edition, Feb 14, 2005 7.7 Silberschatz, Galvin and Gagne ©2005
 Resource-Allocation Graph

A set of vertices V and a set of edges E.

 V is partitioned into two types:
 P = {P1, P2, …, Pn}, the set consisting of all the processes
in the system.

 R = {R1, R2, …, Rm}, the set consisting of all resource types

in the system.
 request edge – directed edge P1  Rj

 assignment edge – directed edge Rj  Pi

Operating System Concepts - 7th Edition, Feb 14, 2005 7.8 Silberschatz, Galvin and Gagne ©2005
 Resource-Allocation Graph (Cont.)

 Process

 Resource Type with 4 instances

 Pi requests instance of Rj
Pi
Rj

 Pi is holding an instance of Rj
Pi
Rj

Operating System Concepts - 7th Edition, Feb 14, 2005 7.9 Silberschatz, Galvin and Gagne ©2005
 Example of a Resource Allocation Graph
(no deadlock) (p3 could release R3)

Operating System Concepts - 7th Edition, Feb 14, 2005 7.10 Silberschatz, Galvin and Gagne ©2005
 Resource Allocation Graph With A Deadlock

Operating System Concepts - 7th Edition, Feb 14, 2005 7.11 Silberschatz, Galvin and Gagne ©2005
 Graph With A Cycle But No Deadlock
because P2 and P4 could release the instances of R1 and
R2; then P1 will use R1 and P3 will use R2 (no dead lock)

Operating System Concepts - 7th Edition, Feb 14, 2005 7.12 Silberschatz, Galvin and Gagne ©2005
 Graph With A Cycle But No Deadlock
because R1 have two instances and one of them is free;
then P1 will use R1 instance and it have R2 already to
finish its execution (no dead lock)

Operating System Concepts - 7th Edition, Feb 14, 2005 7.13 Silberschatz, Galvin and Gagne ©2005
 Basic Facts

 If graph contains no cycles  no deadlock.

 If graph contains a cycle 

 if only one instance per resource type, then
deadlock.
 if several instances per resource type,
possibility of deadlock.

Operating System Concepts - 7th Edition, Feb 14, 2005 7.14 Silberschatz, Galvin and Gagne ©2005
 Methods for Handling Deadlocks

1) Ensure that the system will never enter a deadlock state

(prevention\avoidance).

2) Allow the system to enter a deadlock state and then recover

(detection then recovery).

3) Ignore the problem and pretend that deadlocks never occur in

the system; used by most operating systems, including UNIX
(ignoring).

Operating System Concepts - 7th Edition, Feb 14, 2005 7.15 Silberschatz, Galvin and Gagne ©2005
 Deadlock Prevention
Prevent at least one condition from the conditions that must
be hold for the deadlock to occur. The possible side effects for this method (prevention)
are: low device utilization and the reduction of system throughput.
 Mutual Exclusion – it is not possible to prevent deadlock by denying this
condition because:
 Mutual exclusion is not required for sharable resources (like read-only
files; these files can be shared with no problems).
 Mutual exclusion must hold for nonsharable resources (like printers, Such
resources are instinctively non-sharable ).

 Hold and Wait – must guarantee- using a protocol- that whenever a process
requests a resource, it does not hold any other resources.
 Require process to request and be allocated all its resources before it
begins execution, or allow process to request resources only when the
process has none.
 This have disadvantages such as: low resource utilization and
starvation.
– Starvation could happen for processes that needs many recourses,
in this case this process may wait for long time periods until all its
requested resources are available together.

Operating System Concepts - 7th Edition, Feb 14, 2005 7.16 Silberschatz, Galvin and Gagne ©2005
 Deadlock Prevention (Cont.)

 No Preemption – to prevent this condition from happening you

can allow the held resource to be preempted.
 If a process that is holding some resources requests another
resource that cannot be immediately allocated to it, then all
resources currently being held are released.
 Preempted resources are added to the list of resources for
which the process is waiting.
 Process will be restarted only when it can regain its old
resources, as well as the new ones that it is requesting.

Operating System Concepts - 7th Edition, Feb 14, 2005 7.17 Silberschatz, Galvin and Gagne ©2005
 Deadlock Prevention (Cont.)
 Circular Wait – to prevent this condition from happening we can impose
a total ordering of all resource types, and require that each process
requests resources in an increasing order of enumeration.
 Ex: we will give each resource a number, such as
 tape drive =1 , disk drive =2, printer =3.
 To prevent the circular wait condition implement the following protocol
which composed from two conditions:
 If a process A needs two resources (tape drive and disk drive) then it
should order the tape drive first, then it order the disk drive.
 If a process A holds the printer and needs the tape drive, it must first
release the printer, then order the tape drive, after this it can order
the printer.
 The previous protocol can be proved (not required), for more details see
book page 255.

Operating System Concepts - 7th Edition, Feb 14, 2005 7.18 Silberschatz, Galvin and Gagne ©2005
 Deadlock Avoidance
Requires that the system has some additional a priori information --to help in
Avoiding the dead lock from happening-- to be available, such information may
Include the previous knowledge about a process Q which will need the tape drive then
the printer. Other process W will need the Printer then the tape drive, and etc……

 Many algorithms for avoidance are available; each of them differ in

the amount and the type of information required (the priori
information). The simplest and most useful model for deadlock
avoidance requires that each process declare the maximum number
of resources of each type that it may need.
 The deadlock-avoidance algorithm dynamically examines the
resource-allocation state ( which have the priori information that are
available, these information includes what resources are needed and
how the process will order and use them) to ensure that there can
never be a circular-wait condition.
 Resource-allocation state is defined by the number of available and
allocated resources, and the maximum demands of the processes.
Operating System Concepts - 7th Edition, Feb 14, 2005 7.19 Silberschatz, Galvin and Gagne ©2005
 Safe State : it is an algorithm for deadlock
avoidance
 When a process requests an available resource, system must decide if
immediate allocation leaves the system in a safe state.
 System is in safe state if there exists a sequence <P1, P2, …, Pn> of ALL the
processes in the systems such that for each P i, the resources that Pi can
still request can be satisfied by currently available resources + resources
held by all the Pj, with j < i.
 The sequence should contains ALL the processes in the system, the
algorithm organizes them in an increasing order (just give them numbers
starting from 0 or 1).
 Ex: If a process (P3) needs 5 resources, two of them are free and
available, the remaining three resources are held by P1 and P2 (P1,and
P2 are lower than P3), then P3 can request those resources from P1 and
P2 (only when they are finished from using those resources) to execute
its job. But if the three resources are held by P6 and P7 then P3 can not
request them unless they are available or held by a lower Processes
(P1,P0,and P2).

Operating System Concepts - 7th Edition, Feb 14, 2005 7.20 Silberschatz, Galvin and Gagne ©2005
 Safe State : it is an algorithm for deadlock
avoidance
 System is in safe state if there exists a sequence <P1, P2, …, Pn> of
ALL the processes in the systems such that for each Pi, the
resources that Pi can still request can be satisfied by currently available
resources + resources held by all the Pj, with j < i.
 That is:
 If Pi resource needs are not immediately available, then Pi can wait
until all Pj have finished.
 When Pj is finished, Pi can obtain needed resources, execute,
return allocated resources, and terminate.
 When Pi terminates, Pi +1 can obtain its needed resources, and so
on.
 Disadvantage: low resource utilization because we have waiting
processes.

Operating System Concepts - 7th Edition, Feb 14, 2005 7.21 Silberschatz, Galvin and Gagne ©2005
 Basic Facts

 If a system is in safe state  no deadlocks.

 If a system is in unsafe state  possibility of

deadlock.

 Avoidance  ensure that a system will never enter

an unsafe state.

Operating System Concepts - 7th Edition, Feb 14, 2005 7.22 Silberschatz, Galvin and Gagne ©2005
 Safe, Unsafe , Deadlock State

possibility of
deadlock.
Not guaranteed
to enter a
deadlock if you
are in the
unsafe area

Operating System Concepts - 7th Edition, Feb 14, 2005 7.23 Silberschatz, Galvin and Gagne ©2005
 Avoidance algorithms

 Single instance of a resource type. Use a resource-

allocation graph

 Multiple instances of a resource type. Use the

banker’s algorithm

Operating System Concepts - 7th Edition, Feb 14, 2005 7.24 Silberschatz, Galvin and Gagne ©2005
 Resource-Allocation Graph Scheme

 Claim edge Pi - - - -> Rj indicated that process Pj may request resource

Rj; represented by a dashed line.

 Claim edge converts to request edge when a process requests a

resource. Ex: Pi  Rj
 Request edge converted to an assignment edge when the resource is
allocated to the process. Ex: Rj  Pi

 When a resource is released by a process, assignment edge reconverts

to a claim edge.
(dashed line)
 Resources must be claimed a priori (priori means that the needed
resources for each process are in advanced known in terms of what and
how) in the system.

Operating System Concepts - 7th Edition, Feb 14, 2005 7.25 Silberschatz, Galvin and Gagne ©2005
 Resource-Allocation Graph
Case A: if we assign R2 to P1 then no circle and no dead lock
Case B: if we assign R2 to P2 then we will have circle and a deadlock

Operating System Concepts - 7th Edition, Feb 14, 2005 7.26 Silberschatz, Galvin and Gagne ©2005
 Unsafe State (circle exists in the graph) In
Resource-Allocation Graph

Operating System Concepts - 7th Edition, Feb 14, 2005 7.27 Silberschatz, Galvin and Gagne ©2005
 Resource-Allocation Graph Algorithm

 Suppose that process Pi requests a resource Rj

 The request can be granted only if converting

the request edge to an assignment edge does
not result in the formation of a cycle in the
resource allocation graph

Operating System Concepts - 7th Edition, Feb 14, 2005 7.28 Silberschatz, Galvin and Gagne ©2005
 Banker’s Algorithm

 Its idea can be implemented in a bank in which the bank never

allocated its available cash in such a way that it could no
longer satisfy the needs of all its customers.
 Multiple instances from each resource are available.

 Each process must a priori claim maximum number of each

resource that it may use.

 When a process requests a resource it may have to wait.

 When a process gets all its resources it must return them in a

finite amount of time.

Operating System Concepts - 7th Edition, Feb 14, 2005 7.29 Silberschatz, Galvin and Gagne ©2005
 Data Structures for the Banker’s Algorithm

Let n = number of processes, and m = number of resources types.

 Available: Vector of length m. If available [j] = k,
there are k instances of resource type Rj available.
 Max: n x m matrix. If Max [i,j] = k, then process Pi
may request at most k instances of resource type Rj.
 Allocation: n x m matrix. If Allocation[i,j] = k then Pi
is currently allocated k instances of Rj.
 Need: n x m matrix. If Need[i,j] = k, then Pi may need
k more instances of Rj to complete its task.

Need [i,j] = Max[i,j] – Allocation [i,j].

Operating System Concepts - 7th Edition, Feb 14, 2005 7.30 Silberschatz, Galvin and Gagne ©2005
 Safety Algorithm

1. Let Work and Finish be vectors of length m and n,

respectively. Initialize:
Work = Available
Finish [i] = false for i = 0, 1, …, n- 1.
2. Find and i such that both:
(a) Finish [i] = false
(b) Needi  Work
If no such i exists, go to step 4.
3. Work = Work + Allocationi
Finish[i] = true
go to step 2.
4. If Finish [i] == true for all i, then the system is in a safe state.

Operating System Concepts - 7th Edition, Feb 14, 2005 7.31 Silberschatz, Galvin and Gagne ©2005
 Resource-Request Algorithm for Process Pi

Request = request vector for process Pi. If Requesti [j] = k

then process Pi wants k instances of resource type Rj.
1. If Requesti  Needi go to step 2. Otherwise, raise error
condition, since process has exceeded its maximum claim.
2. If Requesti  Available, go to step 3. Otherwise Pi must
wait, since resources are not available.
3. Pretend to allocate requested resources to Pi by modifying
the state as follows:
Available = Available – Request;
Allocationi = Allocationi + Requesti;
Needi = Needi – Requesti;
 If safe  the resources are allocated to Pi.
 If unsafe  Pi must wait, and the old resource-allocation
state is restored

Operating System Concepts - 7th Edition, Feb 14, 2005 7.32 Silberschatz, Galvin and Gagne ©2005
 Example of Banker’s Algorithm

 5 processes P0 through P4;

3 resource types:
A (10 instances), B (5instances), and C (7 instances).
 Snapshot at time T0:

Allocation Max Available

A B CA B C ABC
P0 010 753 332
P1 200 322
P2 302 902
P3 211 222
P4 002 433

Operating System Concepts - 7th Edition, Feb 14, 2005 7.33 Silberschatz, Galvin and Gagne ©2005
 Example (Cont.)

 The content of the matrix Need is defined to be Max – Allocation.

Need
ABC
P0 743
P1 122
P2 600
P3 011
P4 431
 The system is in a safe state since the sequence < P1, P3, P4, P2,
P0> satisfies safety criteria.
 can you create another sequence that satisfies the safe state? yes

Operating System Concepts - 7th Edition, Feb 14, 2005 7.34 Silberschatz, Galvin and Gagne ©2005
 Example: P1 Request (1,0,2)
 Check that Request  Available (that is, (1,0,2)  (3,3,2)  true.

Allocation Need Available

ABC ABC ABC
P0 010 743 230
P1 302 020
P2 301 600
P3 211 011
P4 002 431
 Executing safety algorithm shows that sequence < P1, P3, P4, P0, P2> satisfies
safety requirement.
 Can request for (3,3,0) by P4 be granted? No because the resources are not
available.
 Can request for (0,2,0) by P0 be granted? No; even though the resources are
available; because this will leave the system in unsafe state (we always consider
the Need (7,4,3) not the request (0,2,0)).
Operating System Concepts - 7th Edition, Feb 14, 2005 7.35 Silberschatz, Galvin and Gagne ©2005
 Deadlock Detection

 If the OS did not prevent or avoid the deadlock then it is

expected to have a deadlock.
 Allow system to enter deadlock state then we should have:

 Detection algorithm

 Recovery scheme

Operating System Concepts - 7th Edition, Feb 14, 2005 7.36 Silberschatz, Galvin and Gagne ©2005
 Single Instance of Each Resource Type

 Maintain wait-for graph to detect the deadlock.

 Wait-for graph is a variant of the resource allocation graph.
 Nodes are processes.
 Pi  Pj if Pi is waiting for Pj.

 Periodically invoke an algorithm that searches for a cycle in the

graph. If there is a cycle, there exists a deadlock.

Operating System Concepts - 7th Edition, Feb 14, 2005 7.37 Silberschatz, Galvin and Gagne ©2005
 Resource-Allocation Graph and Wait-for Graph

Resource-Allocation Graph Corresponding wait-for graph

Operating System Concepts - 7th Edition, Feb 14, 2005 7.38 Silberschatz, Galvin and Gagne ©2005
 Several Instances of a Resource Type

 Available: A vector of length m indicates the number

of available resources of each type.

 Allocation: An n x m matrix defines the number of

resources of each type currently allocated to each
process.

 Request: An n x m matrix indicates the current

request of each process. If Request [ij] = k, then
process Pi is requesting k more instances of resource
type. Rj.

Operating System Concepts - 7th Edition, Feb 14, 2005 7.39 Silberschatz, Galvin and Gagne ©2005
 Detection Algorithm

1. Let Work and Finish be vectors of length m and n,

respectively Initialize:
(a) Work = Available
(b) For i = 1,2, …, n, if Allocationi  0, then
Finish[i] = false;otherwise, Finish[i] = true.
2. Find an index i such that both:
(a) Finish[i] == false
(b) Requesti  Work

If no such i exists, go to step 4.

Operating System Concepts - 7th Edition, Feb 14, 2005 7.40 Silberschatz, Galvin and Gagne ©2005
 Detection Algorithm (Cont.)

3. Work = Work + Allocationi

Finish[i] = true
go to step 2.

4. If Finish[i] == false, for some i, 1  i  n, then the system is in

deadlock state. Moreover, if Finish[i] == false, then Pi is
deadlocked.

Algorithm requires an order of O(m x n2) operations to detect

whether the system is in deadlocked state.

Operating System Concepts - 7th Edition, Feb 14, 2005 7.41 Silberschatz, Galvin and Gagne ©2005
 Example of Detection Algorithm

 Five processes P0 through P4; three resource types

A (7 instances), B (2 instances), and C (6 instances).
 Snapshot at time T0:

Allocation Request Available

ABC ABC ABC
P0 010 000 000
P1 200 202
P2 303 000
P3 211 100
P4 002 002
 Sequence <P0, P2, P3, P1, P4> will result in Finish[i] = true for all i.

Operating System Concepts - 7th Edition, Feb 14, 2005 7.42 Silberschatz, Galvin and Gagne ©2005
 Example (Cont.)

 If P2 requests an additional instance of type C.

Request
ABC
P0 000
P1 202
P2 001
P3 100
P4 002
 State of system?
 Can reclaim resources held by process P0, but insufficient
resources to fulfill other processes; requests.
 Deadlock exists, consisting of processes P1, P2, P3, and P4.

Operating System Concepts - 7th Edition, Feb 14, 2005 7.43 Silberschatz, Galvin and Gagne ©2005
 Detection-Algorithm Usage
 When, and how often, to invoke the detection algorithm? The answer
depends on two factors:
 How often a deadlock is likely to occur?
 How many processes will need to be rolled back?
 If we invoke the detection algorithm every time a request for allocation of
a resource can not be granted, then we can find the process which
cause the dead lock to occur directly (it will be the last one so we role it
back).
 Disadvantage: the cost for using the detection algorithm frequently is
high.
 If detection algorithm is invoked in intervals (ex: every 60 seconds) or
when the CPU utilization drops under 40 percent then we will have the
following disadvantage:
 there may be many cycles in the resource graph and so we
would not be able to tell which of the many deadlocked processes
“caused” the deadlock so we have –maybe- to role back them all.

Operating System Concepts - 7th Edition, Feb 14, 2005 7.44 Silberschatz, Galvin and Gagne ©2005
 Recovery from Deadlock: Process Termination
 To recover from the deadlock we have three main methods:
 Process termination: we have two options:
 Abort all deadlocked processes.
 Abort one process at a time until the deadlock cycle is eliminated.
 Tell the operator (user) to deal with it manually. (ex: restart the
computer).
 Resource preemption: we successively preempt some resources from
processes and give these resources to other processes until the
deadlock is terminated.
 In which order should we choose to abort? It depends on:
 Priority of the process.
 How long process has computed, and how much longer to completion.
 Resources the process has used.
 Resources process needs to complete.
 How many processes will need to be terminated.
 Is process interactive or batch? 7.45
Operating System Concepts - 7th Edition, Feb 14, 2005 Silberschatz, Galvin and Gagne ©2005
 Recovery from Deadlock: Resource Preemption

 If preemption required to deal with deadlocks, then three issues need to

be addressed:
 Selecting a victim – which process and resource we should choose
to terminate? Always try to minimize cost based on same factors
mentioned in the previous slide.
 Rollback – return to some safe state, restart process for that state.
 Starvation – how do we guarantee that the same process may not
always be picked as victim when the deadlock occurs? If the same
process always picked as a victim then starvation will occur for this
process.
 You can use a rollback strategy or the victim selection strategy to
preempt resources, each of which have its own advantages and
disadvantages.

Operating System Concepts - 7th Edition, Feb 14, 2005 7.46 Silberschatz, Galvin and Gagne ©2005
End of Chapter 7