1. Chapter 3 Deadlocks - Αδιέξοδα 3.1. Resource
3.2. Introduction to deadlocks
3.3. The ostrich algorithm
3.4. Deadlock detection and recovery
3.5. Deadlock avoidance
3.6. Deadlock prevention
3.7. Other issues

2. Resources - Πόροι Examples of computer resources
printers
tape drives
tables
Processes need access to resources in reasonable order
Suppose a process holds resource A and requests resource B
at same time another process holds B and requests A
both are blocked and remain so
Hardware and software deadlocks

3. Resources Deadlocks occur when …
processes are granted exclusive access to devices
we refer to these devices generally as resources
Resources may have multiple copies
Preemptable (προεκχωρήσιμοι) resources
can be taken away from a process with no ill effects (for example memory)
Nonpreemptable (μη-προεκχωρήσιμοι) resources
will cause the process to fail if taken away (e.g. CDR)

4. Resources Sequence of events required to use a resource
request the resource
use the resource
release the resource
Must wait if request is denied
requesting process may be blocked
may fail with error code
Nature of requesting a resource is highly system dependent (e.g. request system call)

5. Resource Acquisition t

6. Introduction to Deadlocks
Formal definition : A set of processes is deadlocked if each process in the set is waiting for an event that only another process in the set can cause
Only one thread, no interrupts
Usually the event is release of a currently held resource
None of the processes can …
run
release resources
be awakened
Number of processes and resources is unimportant

7. Four Conditions for Deadlock
Mutual exclusion condition
each resource assigned to 1 process or is available
Hold and wait condition
process holding resources can request additional
No preemption condition
previously granted resources cannot forcibly taken away
Circular wait condition
must be a circular chain of 2 or more processes
each is waiting for resource held by next member of the chain
All relate to a policy that a system can or can not have

8. Deadlock Modeling Modeled with directed graphs
resource R assigned to process A
process B is requesting/waiting for resource S
process C and D are in deadlock over resources T and U

9. Deadlock Modeling
A B C
How deadlock occurs

10. How deadlock can be avoided
Deadlock Modeling
(o) (p) (q)
How deadlock can be avoided

11. Deadlock Modeling Strategies for dealing with Deadlocks
just ignore the problem altogether
detection and recovery (ανίχνευση και επανόρθωση)
dynamic avoidance (αποφυγή)
careful resource allocation
prevention (πρόληψη)
negating one of the four necessary conditions

12. The Ostrich Algorithm – Αλγ. στρουθοκαμήλου
Pretend there is no problem
Reasonable if
deadlocks occur very rarely
cost of prevention is high
UNIX and Windows takes this approach
It is a trade off between
convenience
correctness

13. Detection with One Resource of Each Type
A…G: processes; R…W: resources
Note the resource ownership and requests
Is this system deadlocked and if yes, which processes are involved?
A cycle can be found within the graph, denoting deadlock

14. Detection with One Resource of Each Type
We need a formal algorithm for detecting deadlocks
A simple one to detect cycles:
Take each node in turn.
Do a DFS (depth first search) on it.
If it comes to a node it has encountered in this run, then there exists a cycle.
Previous graph has a cycle

15. Detection with Multiple Resources of Each Type
Data structures needed by deadlock detection algorithm
At all times: Σi=1Cij + Aj = Ej n

16. Detection with Multiple Resources of Each Type
Deadlock detection is based on comparing vectors:
Algorithm
Look for an unmarked process, Pi for which the i-th row of R is less or equal to A
If such a process is found, add the i-th row of C to A, mark the process and go back to step 1
If no such process exists the algorithm terminates

17. Detection with Multiple Resources of Each Type
An example for the deadlock detection algorithm
(3/2/1)

18. Detection with Multiple Resources of Each Type
When to look for deadlocks?
Every time a resource request is made
Detection ASAP
Expensive
Every k minutes or whenever the CPU utilization drops below a certain threshold

19. Recovery from Deadlock - Επανόρθωση
Recovery through preemption
take a resource from some other process (e.g. printer)
depends on nature of the resource
Recovery through rollback
checkpoint a process periodically
use this saved state
restart the process if it is found deadlocked

20. Recovery from Deadlock
Recovery through killing processes
crudest but simplest way to break a deadlock
kill one of the processes in the deadlock cycle
the other processes get its resources
choose process that can be rerun from the beginning (perhaps not in cycle)

21. Deadlock Avoidance - Αποφυγή
So far we assumed that all requests take place at the beginning
The system must be able to decide whether granting a resource request is safe or not
Is there an algorithm that can always avoid deadlocks?
Yes, if certain information is known in advance

22. Deadlock Avoidance - Resource Trajectories
Two process resource trajectories
/// and \\\ are impossible to get
What scheduler should do at point t ?

23. Demonstration that the state in (a) is safe – 10 instances
Safe and Unsafe States
A state is said to be safe if it is not deadlocked and there is some scheduling order in which every process can run to completion even if all of them suddenly request their maximum number of resources immediately
(a) (b) (c) (d) (e)
Demonstration that the state in (a) is safe – 10 instances

24. Safe and Unsafe States Demonstration that the state in b is not safe
(a) (b) (c) (d)
Demonstration that the state in b is not safe
An unsafe state is not a deadlocked state

25. The Banker's Algorithm for a Single Resource
(a) (b) (c)
Check to see if granting the request leads to unsafe state
Three resource allocation states
safe
unsafe

26. Banker's Algorithm for Multiple Resources
Example of banker's algorithm with multiple resources

27. Banker's Algorithm for Multiple Resources
Look for a row, R, whose unmet resource needs are all smaller than or equal to A. If no such row exists the system will eventually deadlock since no process can run to completion.
Assume the process of the row chosen requests all the resources it needs (which is guaranteed to be possible) and finishes. Mark that process as terminated and add all its resources to the A vector
Repeat step 1 and 2 until either all processes are marked terminated, in which case the state is safe, or until a deadlock occurs, in which case is not.
B requests a printer… (D,A or E, …)
E requests a printer… (deadlock)

28. Deadlock Prevention Attacking the Mutual Exclusion Condition
Some devices (such as printer) can be spooled
only the printer daemon uses printer resource
thus deadlock for printer eliminated
Not all devices can be spooled
Principle:
avoid assigning resource when not absolutely necessary
as few processes as possible actually claim the resource

29. Attacking the Hold and Wait Condition
Goal: Prevent processes that hold resources from waiting for more resources
Require processes to request resources before starting
a process never has to wait for what it needs
Problems
may not know required resources at start of run
also ties up resources other processes could be using
Variation:
process must give up temporarily all resources before requesting a new one
then request all immediately needed

30. Attacking the No Preemption Condition
This is not a viable option
Consider a process given the printer
halfway through its job
now forcibly take away printer
!!??

31. Attacking the Circular Wait Condition
(a) (b)
Normally ordered resources
A resource graph

32. Attacking the Circular Wait Condition
Rule: All requests of a process must be made in numerical order => the resource allocation graph can not have cycles
Either i < j or i > j => can’t have deadlocks
Same logic with multiple resources: at every instant one assigned resource will be the highest
Problem: impossible to find an ordering to satisfy everyone

33. Attacking the Circular Wait Condition
Summary of approaches to deadlock prevention
Avoidance and prevention are not widely used in OS, but have special-purpose applications

34. Other Issues Two-Phase Locking
DB systems lock records for update
Phase One
process tries to lock all records it needs, one at a time
if needed record found locked, start over
(no real work done in phase one)
If phase one succeeds, it starts second phase,
performing updates
releasing locks
Note similarity to requesting all resources at once
Algorithm works where programmer can arrange things so that the program can be stopped and restarted

35. Non-resource Deadlocks
Possible for two processes to deadlock
each is waiting for the other to do some task
Can happen with semaphores
each process required to do a down() on two semaphores (mutex and another)
if done in wrong order, deadlock results

36. Starvation Algorithm to allocate a resource
may be to give to shortest job first
Works great for multiple short jobs in a system
May cause long job to be postponed indefinitely
even though not blocked
Solution:
First-come, first-serve policy