Chapter 8: Deadlocks

Operating System Concepts – 10 th Edition Silberschatz, Galvin and Gagne

 Chapter 8: Deadlocks
 System Model
 Deadlock Characterization
 Methods for Handling Deadlocks
 Deadlock Prevention
 Deadlock Avoidance
 Deadlock Detection
 Recovery from Deadlock

Operating System Concepts – 10 th Edition 8.2 Silberschatz, Galvin and Gagne

 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 – 10 th Edition 8.3 Silberschatz, Galvin and Gagne

 Definitions

 Threads – active
 Resources – passive, things needed by thread to do its job
 CPU, disk space, memory

 Two kinds of resources:

 Preemptable – can take it away
 CPU, Embedded security chip
 Non-preemptable – must leave with thread
 disk space, plotter, chunk of virtual address space
 Mutual exclusion – the right to enter a critical section is a
kind of resource

Operating System Concepts – 10 th Edition 8.4 Silberschatz, Galvin and Gagne

 Resources
 Resources may require exclusive access or may be sharable
 Read-only files are typically sharable
 Printers are not sharable during time of printing

 One of the major tasks of an operating system is to manage

resources

Operating System Concepts – 10 th Edition 8.5 Silberschatz, Galvin and Gagne

 Starvation vs Deadlock
 Starvation – thread waits indefinitely
 Example, low-priority thread waiting for resources
constantly in use by high-priority threads
 Deadlock – circular waiting for resources
 Example:
 I have resource A and need resource B to get my job
done.
 You have resource B and need resource A to get your
job done.
 Deadlock implies starvation, but not vice versa
 Starvation can end (but doesn’t have to)
 Deadlock can’t end without external intervention.

Operating System Concepts – 10 th Edition 8.6 Silberschatz, Galvin and Gagne

 Conditions for deadlock - Motivation
 Deadlock need not be deterministic:
 semaphores A and B, initialized to 1

P0 P1
wait (A); wait(B);
wait (B); wait(A);
Signal(B); signal(A);
Signal(A); Signal(B);

 Deadlock won’t always happen with this code, but it might.

 Have to have exactly the right timing (“wrong” timing?)
 So you release a piece of software, and you tested it,
and there it is, controlling a nuclear power plant

Operating System Concepts – 10 th Edition 8.7 Silberschatz, Galvin and Gagne

 Conditions for deadlock - Motivation
 Deadlocks can occur with multiple resources.
 Means you can’t decompose the problem
 can’t solve deadlock for each resource independently.
 For example:
1. One thread grabs the memory it needs
2. Another grabs disk space
3. Another grabs the tape drive
 Each waits for the other to release.

 Deadlock can occur whenever there is waiting.

 Example: dining philosophers
 Each philosopher needs two chopsticks to eat. Each grabs
chopstick on the right first.
 What if all grab at the same time? Deadlock.

Operating System Concepts – 10 th Edition 8.8 Silberschatz, Galvin and Gagne

 System Model
 System consists of resources
 Resource types R1, R2, . . ., Rm
CPU cycles, memory space, I/O devices
 Each resource type Ri has Wi instances.
 Each process utilizes a resource as follows:
 request
 use
 release

Operating System Concepts – 10 th Edition 8.9 Silberschatz, Galvin and Gagne

 Deadlock Characterization
Deadlock can arise if four conditions hold
simultaneously.
 Mutual exclusion: only one process at a time can use a
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 {P0, P1, …, Pn} of waiting
processes 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 Pn
is waiting for a resource that is held by P0.

Operating System Concepts – 10 th Edition 8.10 Silberschatz, Galvin and Gagne

 Deadlock with Mutex Locks
 Deadlocks can occur via system calls, locking, etc.
 See example box in text page 318 for mutex deadlock

Operating System Concepts – 10 th Edition 8.11 Silberschatz, Galvin and Gagne

 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 Pi  Rj

 assignment edge – directed edge Rj  Pi

Operating System Concepts – 10 th Edition 8.12 Silberschatz, Galvin and Gagne

 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 – 10 th Edition 8.13 Silberschatz, Galvin and Gagne

 Resource Allocation Graph Examples

Simple Resource Allocation Graph Allocation Graph With

Allocation Graph With Deadlock Cycle, but No Deadlock
Operating System Concepts – 10 th Edition 8.14 Silberschatz, Galvin and Gagne
 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 – 10 th Edition 8.15 Silberschatz, Galvin and Gagne

 Methods for Handling Deadlocks
 Ensure that the system will never enter a deadlock
state:
 Deadlock prevention
 Deadlock avoidance
 Allow the system to enter a deadlock state and then
recover
 Ignore the problem and pretend that deadlocks never
occur in the system; used by most operating systems,
including UNIX, JVM

 Deadlock prevention - eliminate one of the necessary

conditions

 Deadlock avoidance – list resources required by a

process in advance

Operating System Concepts – 10 th Edition 8.16 Silberschatz, Galvin and Gagne

 Deadlock Prevention
Restrain the ways request can be made

 Mutual Exclusion – not required for sharable resources (e.g.,

read-only files); must hold for non-sharable resources
 Buy more resources, split into pieces, or virtualize to
make “infinite” copies
 Hold and Wait – must guarantee 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 allocated to it.
 Low resource utilization; starvation possible

Operating System Concepts – 10 th Edition 8.17 Silberschatz, Galvin and Gagne

 Deadlock Prevention (Cont.)
 No Preemption –
 make copies and preempt
 Can preempt main memory by copying to disk
 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
 Circular Wait – impose a total ordering of all resource types,
and require that each process requests resources in an
increasing order of enumeration

Operating System Concepts – 10 th Edition 8.18 Silberschatz, Galvin and Gagne

 Deadlock Example
/* thread one runs in this function */
void *do_work_one(void *param)
{

pthread_mutex_lock(&first_mutex);
pthread_mutex_lock(&second_mutex);
/** * Do some work */
pthread_mutex_unlock(&second_mutex);
pthread_mutex_unlock(&first_mutex);
pthread_exit(0);
}
/* thread two runs in this function */
void *do_work_two(void *param)
{

pthread_mutex_lock(&second_mutex);
pthread_mutex_lock(&first_mutex);
/** * Do some work */
pthread_mutex_unlock(&first_mutex);
pthread_mutex_unlock(&second_mutex);
pthread_exit(0);
}

Operating System Concepts – 10 th Edition 8.19 Silberschatz, Galvin and Gagne

 Deadlock Example with Lock Ordering
void transaction(Account from, Account to, double amount)
{
mutex lock1, lock2;
lock1 = get_lock(from);
lock2 = get_lock(to);
acquire(lock1);
acquire(lock2);
withdraw(from, amount);
deposit(to, amount);
release(lock2);
release(lock1);
}

Transactions 1 and 2 execute concurrently. Transaction 1

transfers $25 from account A to account B, and Transaction
2 transfers $50 from account B to account A

Operating System Concepts – 10 th Edition 8.20 Silberschatz, Galvin and Gagne

 Deadlock Avoidance
Requires that the system has some additional a priori
information
available
 Simplest and most useful model 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 to ensure that there can never be a
circular-wait condition
 Resource-allocation state: number of available and allocated
resources, and the maximum demands of the processes

Operating System Concepts – 10 th Edition 8.21 Silberschatz, Galvin and Gagne

 Safe State
 When a process requests an available resource, system must decide
if immediate allocation leaves the system in a safe state
 Safe state: if the system can allocate resources to each process (up
to its maximum) in some order and still avoid a deadlock.
 System is in safe state if there exists a safe sequence of all
processes.
 Safe sequence: 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

Operating System Concepts – 10 th Edition 8.22 Silberschatz, Galvin and Gagne

 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 – 10 th Edition 8.23 Silberschatz, Galvin and Gagne

 Safe, Unsafe, Deadlock State

Operating System Concepts – 10 th Edition 8.24 Silberschatz, Galvin and Gagne

 Deadlock Avoidance
 Example
 available - 12 magnetic tape drives, 3 processes: P 0, P1, P2. At time
t0
Maximum Needs Current Needs
P0 10 5
P1 4 2
P2 9 2
 At time t0, the system is in a safe state.
 The sequence <P1,P0,P2> is safe.
 It is possible to go from a safe state to an unsafe state.
 Suppose at time t1, process P2 requests and is allocated 1 more tape
drive.
 The system is no longer in a safe state.
 Avoidance  ensure that a system will never enter an unsafe state.
 Wait if allocation will lead to unsafe state.
Operating System Concepts – 10 th Edition 8.25 Silberschatz, Galvin and Gagne
 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 – 10 th Edition 8.26 Silberschatz, Galvin and Gagne

 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
 Request edge converted to an assignment edge when the
resource is allocated to the process
 When a resource is released by a process, assignment edge
reconverts to a claim edge
 Resources must be claimed a priori in the system

Operating System Concepts – 10 th Edition 8.27 Silberschatz, Galvin and Gagne

 Resource-Allocation Graph

Operating System Concepts – 10 th Edition 8.28 Silberschatz, Galvin and Gagne

 Unsafe State In Resource-Allocation Graph

Operating System Concepts – 10 th Edition 8.29 Silberschatz, Galvin and Gagne

 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 – 10 th Edition 8.30 Silberschatz, Galvin and Gagne

 Banker’s Algorithm
 Multiple instances

 Each process must a priori claim maximum 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 – 10 th Edition 8.31 Silberschatz, Galvin and Gagne

 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 – 10 th Edition 8.32 Silberschatz, Galvin and Gagne

 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 an 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

 This algorithm may require an order of m x n2

operations to decide whether a state is safe.
Operating System Concepts – 10 th Edition 8.33 Silberschatz, Galvin and Gagne
 Resource-Request Algorithm for Process Pi

Requesti = 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 – Requesti;
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 – 10 th Edition 8.34 Silberschatz, Galvin and Gagne

 Example of Banker’s Algorithm

 5 processes P0 through P4;

3 resource types:
A (10 instances), B (5 instances), and C (7
instances)
 Snapshot at time T0:
Allocation Max Available
ABC ABC ABC
P0 010 753 332
P1 200 322
P2 302 902
P3 211 222
P4 002 433

Operating System Concepts – 10 th Edition 8.35 Silberschatz, Galvin and Gagne

 Example (Cont.)
 The content of the matrix Need is defined to be Max –
Allocation

Need Available
ABC ABC
P0 743 332
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

Operating System Concepts – 10 th Edition 8.36 Silberschatz, Galvin and Gagne

 Example: P1 Request (1,0,2)
 Check that Request  Available (that is, (1,0,2)  (3,3,2) 
true
Allocation Need Available
A B CA B C ABC
P0 010 743 230
P1 302 020
P2 302 600
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?

 Can request for (0,2,0) by P0 be granted?

Operating System Concepts – 10 th Edition 8.37 Silberschatz, Galvin and Gagne

 Deadlock Detection

 Allow system to enter deadlock state

 Detection algorithm

 Recovery scheme

 Detection and recovery scheme requires overhead

that includes
 run-time costs of maintaining the necessary
information and executing the detection algorithm,
 the potential losses inherent in recovering from a
deadlock.

Operating System Concepts – 10 th Edition 8.38 Silberschatz, Galvin and Gagne

 Single Instance of Each Resource Type

 Maintain wait-for 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

 An algorithm to detect a cycle in a graph requires an

order of n2 operations, where n is the number of
vertices in the graph

Operating System Concepts – 10 th Edition 8.39 Silberschatz, Galvin and Gagne

 Resource-Allocation Graph and Wait-for Graph

Resource-Allocation Graph
Corresponding wait-for graph

Operating System Concepts – 10 th Edition 8.40 Silberschatz, Galvin and Gagne

 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 [i][j] = k, then process Pi is
requesting k more instances of resource type Rj.

 The detection algorithm investigates every possible

allocation sequence for the processes that remain to be
completed.

Operating System Concepts – 10 th Edition 8.41 Silberschatz, Galvin and Gagne

 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
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 – 10 th Edition 8.42 Silberschatz, Galvin and Gagne

 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 – 10 th Edition 8.43 Silberschatz, Galvin and Gagne

 Example (Cont.)

 P2 requests an additional instance of type C

Request
ABC
P0 000
P1 202
P2 001
P3 1 0 0
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 – 10 th Edition 8.44 Silberschatz, Galvin and Gagne

 Detection-Algorithm Usage
 When, and how often, to invoke depends on:
 How often a deadlock is likely to occur?
 How many processes will need to be rolled back?
 one for each disjoint cycle

 If detection algorithm is invoked arbitrarily, 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.
 Frequent deadlocks  invoke detection algorithm frequently.
 Extreme case: invoke detection algorithm every time a request
for allocation cannot be granted immediately.
 A less expensive alternative: invoke the algorithm at less
frequent intervals – e.g. once per hour, or whenever CPU
utilization drops below 40%.

Operating System Concepts – 10 th Edition 8.45 Silberschatz, Galvin and Gagne

 Recovery from Deadlock: Process Termination

 Abort all deadlocked processes

 Expensive (partial results may have to be recomputed)

 Abort one process at a time until the deadlock cycle is

eliminated
 considerable overhead (rerun deadlock-detection algorithm)

 In which order should we choose to abort?

1. Priority of the process
2. How long process has computed, and how much longer to
completion
3. Resources the process has used
4. Resources process needs to complete
5. How many processes will need to be terminated
6. Is process interactive or batch?

Operating System Concepts – 10 th Edition 8.46 Silberschatz, Galvin and Gagne

 Recovery from Deadlock: Resource Preemption

 Selecting a victim – minimize cost

 Which resources and which processes are to be preempted?
 Cost factors include the following parameters:
 number of resources held by a deadlocked process
 amount of time consumed so far by the deadlocked process
 Rollback – return to some safe state, restart process for that state
 Simpler to rollback completely: abort and restart
 partial rollback requires more information to be recorded for all
running processes.
 Starvation – same process may always be picked as victim,
 pick up a process as a victim only (small) finite number of times.
 include number of rollbacks in the cost factor.

Operating System Concepts – 10 th Edition 8.47 Silberschatz, Galvin and Gagne

 Combined Approach to Deadlock Handling

 Combine the three basic approaches,

 prevention
 avoidance
 detection
allowing the use of the optimal approach for each class of
resources in the system.
 Partition resources into hierarchically ordered classes.
 Use most appropriate technique for handling deadlocks
within each class.

Operating System Concepts – 10 th Edition 8.48 Silberschatz, Galvin and Gagne

 End of Chapter 8

Operating System Concepts – 10 th Edition Silberschatz, Galvin and Gagne