Study with the several resources on Docsity
Earn points by helping other students or get them with a premium plan
Prepare for your exams
Study with the several resources on Docsity
Earn points to download
Earn points by helping other students or get them with a premium plan
An in-depth analysis of deadlocks in computer science, including their definition, causes, approaches to detection and prevention, and special cases. Both reusable and consumable resources, examples of deadlocks, and the impact of deadlocks on systems. It also discusses the importance of deadlocks in real-time applications and their potential consequences.
Typology: Study notes
1 / 28
This page cannot be seen from the preview
Don't miss anything!
CompSci 143a
Spring, 2009
6.1 Deadlocks with Reusable and Consumable Resources6.2 Approaches to the Deadlock Problem6.3 A System Model
6.4 Deadlock Detection
6.5 Recovery from Deadlock6.6 Dynamic Deadlock Avoidance
6.7 Deadlock Prevention
CompSci 143a
Spring, 2009
CompSci 143a
Spring, 2009
p1: ...
p2: ...
open(f1,w);
open(f2,w);
open(f2,w);
open(f1,w);
...
...
Deadlock when executed concurrently p1: if (C) send(p2,m);
p2: ...
while(1){...
while(1){...
recv(p2,m);
recv(p1,m);
send(p2,m);
send(p1,m);
... }
... }
Deadlock when
not true
CompSci 143a
Spring, 2009
-^
-^
requests may starve a 200MB request
CompSci 143a
Spring, 2009
Assumptions:– When a process requests a resource, either the request is
fully granted or the process blocks
Resource graph:^ – Vertices are processes, resources, resource units– Edges (directed) represent requests and allocations of
resources
CompSci 143a
Spring, 2009
Resource graph:
Process = CircleResource = Rectangle with small circles for each unitRequest = Edge from process to resource classAllocation = Edge from resource unit to process
Figure 6-
CompSci 143a
Spring, 2009
CompSci 143a
Spring, 2009
11
2 processes
p
1
p
; 2 resources 2
p
1
and
p
2
both need
1
and
2
p
1
requests
1
before
2
and releases
2
before
p
2
requests
2
before
1
and releases
1
before
Figure 6-3: Transitions by
p
only 1
CompSci 143a
Spring, 2009
Graph Reduction:
Repeat the following
p
p
and all request and allocation edges
Deadlock
Graph not completely reducible.
All reduction sequences lead to the same result.
Figure 6-
CompSci 143a
Spring, 2009
-^
Reduce until
p
is removed or graph irreducible
-^
Current state not deadlocked
Next state
deadlocked only if:
a.
Operation was a request by
p
and
b.
p
is deadlocked in
Try to reduce
by
p
CompSci 143a
Spring, 2009
Special Cases of Detection
Knot
: A set K of nodes such that
Deadlock :
(Remove R2->p1: knot R2,p3,R3,p5)(Reverse edge p1->R1): expedient state
CompSci 143a
Spring, 2009
For single-unit resources, cycle
deadlock
p
must have a request edge to
R
R
must have an allocation edge to
p
R
is not available and thus
p
is blocked
Wait-For Graph (wfg):
Show only processes
p
→
R
→
p
2
by
p
→
p
2
:^
p
1
waits for
p
Figure 6-
CompSci 143a
Spring, 2009
Distributed Approach– Detect cycles using probes.– If process
pi
blocked on
p
, it launches probe j
p
→ i
pj
pi
→
pj
→
p
k^
along all request edges, etc.
pi
, cycle is detected Figure 6-
CompSci 143a
Spring, 2009
