Topic 11 – Context switching

Question 1

in regards to the context block:

this is a record of what state the process is in such as running, ready or blocked, it is used to decide how the process should be treated.

Answer 1

in regards to the context block what is the:

process state

Question 2

in 7 steps describe how

buffer control is possible by using a semaphore

Answer 2

the steps that describe this are:

1. a buffer is empty and has a size of 10 (the semaphore value is 10)
2. a process wishes to deposit an item so makes a SEM_WAIT system call
3. since the semaphore is not 0 the process can deposit the item in the buffer
4. the semaphore value is now decremented to 9
5. steps 2 - 5 will reapeat untill the semaphore reaches 0
6. a process wishes to deposit an item so makes a SEM_WAIT system call
7. the semaphor count is 0 and so the process is put into a waiting state until another process removes an item from the buffer and makes the system call SEM_SIGNAL so that the semaphore can be incremented by 1

note

this example does not take into account an event such as two threads colliding at a buffer address (race condition) for this to be mitigated a seperate semaphore would be needed for mutual exclusion (mutex)

Question 3

the main feature of the processor that the operating system takes advantage of to achieve this is:

the fact that a hardware interrupt will pause the execution of any process then switch to kernel mode and hand control of the CPU to the operating system

the hardware interrupt in question is received from a hardware timer

Answer 3

what is the main feature of the processor that the operating system takes advantage of to achieve

pre-emptive multitasking

Question 4

describe in 5 steps the

stack pointer

during an interrupt without a context switch

Answer 4

this without a context switch

1. an interrupt occurs
2. process A instruction pointer is placed on top the stack pointer
3. the routine is placed inside the instruction pointer and is executed
4. when the routine is finished it will call the RETURN instruction
5. process A instruction pointer is pulled of the stack pointer and execution commences where it was paused

Question 5

in regards to the context block:

this records information such as a processes running time (quantum) and other information that can be used by the scheduler so that it can decide the order of execution

Answer 5

in regards to the context block what is the:

process priority

Question 6

what are the two actions that can occur that would

cause a process to be transitioned and control of the CPU handed back to the operating system

Answer 6

these actions include:

1. Transition because of a system call: for example when an application program makes an I/O request or goes to sleep
2. Transition because of an interrupt: for example when the hardware timer interrupts

For either of these the process will change state and the OS will handle the events

Question 7

two use cases of this are:

1. mutual exclusion (mutex) - this can be used by a thread to tell other threads that it is currently executing a critical region of code and that the resource must not be acceessed this avoids scenerios such as a race condition
2. buffer control - this can also be used as a counter for example how many units are available in a buffer

Answer 7

name 2 uses of the

semaphore

Question 8

name 6 pieces of information that would be found in the

context block

for a process

Answer 8

this would record details for a process such as:

• Process Identifier (PID)
• the process state
• the process priority
• saved register contents
• information about the processes memory space
• information about files

Question 9

while the processor is in kernel mode what 2 items does the operating system have access to

Answer 9

while the processor is in this mode the opearting system can access:

• all instructions including privileged instructions
• all data structures including the context block

Question 10

in summary these problems occure when:

1. There is a shared resource
2. There are critical regions of code acting on those shared resources

Answer 10

in two summary points when do Concurrency problems such as the race condition occur

Question 11

this system call essentially asks the OS if it can enter a critical region of code. If the semaphore count is 0 then the process is placed into a waiting state. If the semaphore count is not 0 then the semaphore count is set to 0 and the process enters a critical region of code

Answer 11

describe the semaphore system call

SEM_WAIT

Question 12

with this type of state the operating system knows that the process is currently using the CPU

Answer 12

if a

process is in a running state

what does the operating system know about the process

Question 13

these actions include:

1. Transition because of a system call: for example when an application program makes an I/O request or goes to sleep
2. Transition because of an interrupt: for example when the hardware timer interrupts

For either of these the process will change state and the OS will handle the events

Answer 13

what are the two actions that can occur that would

cause a process to be transitioned and control of the CPU handed back to the operating system

Question 14

this would record details for a process such as:

• Process Identifier (PID)
• the process state
• the process priority
• saved register contents
• information about the processes memory space
• information about files

Answer 14

name 6 pieces of information that would be found in the

context block

for a process

Question 15

if a process is in this state then the operating system knows that it is waiting on an I/O operation to complete

Answer 15

if a

process is in a blocked state

what does the operating system know about the process

Question 16

this can only occur after a system call has been made or a hardware interrupt signal has been received.

in either case the processor will switch to kernel mode and the OS will gain control of the CPU, allowing it a chance to check the quantum of any running processes and run its scheduler if any quantum has expired

Answer 16

when can context switching occur

Question 17

describe the semaphore system call

SEM_SIGNAL

Answer 17

this system call essentially tells the operating system that it is finished executing its critical region of code and the semaphore count can be incremented so that another process may enter its critical region of code.

Question 18

this is a place in the source code that acts on a shared resource. If these actions are not carried out correctly and in full then the value of the shared resource could become false

Answer 18

what is a

critical region

of code

Question 19

this is a solution to multitasking

within this solution the operating system has control over when it will use the CPU.

it does this by setting a hardware timer that will interrupt any process

Answer 19

describe what

pre-emptive multitasking

is

Question 20

this is the term given to a semaphore when it is used for the purpose of mutual exclusion

note

semaphores have many use cases within the operating system

Answer 20

what is a

mutex

Question 21

in regards to the context block what is the:

process priority

Answer 21

in regards to the context block:

this records information such as a processes running time (quantum) and other information that can be used by the scheduler so that it can decide the order of execution

Question 22

this is the act of the currently running process being replaced by another process

note

since all data within the CPU registers will be replaced it is said that the CPU is now executing in a different “context”

Answer 22

what is

context switching

Question 23

if a

process is in a ready state

what does the operating system know about the process

Answer 23

if a process is in this state then the operating system knows that it is waiting to be rescheduled to use the CPU

Question 24

execpt for the semaphore system calls SEM_WAIT and SEM_SIGNAL

what other system calls are there that are associatied with the semaphore

Answer 24

There are also system calls that allow the creation and destroying of different semaphores so that different semaphores can be used by different applications

Question 25

this is a diagram that expresses the transition and states of a process. * ovals show the state of a process * arrows show the direction and transition of a process from one state to another _notes_ * All states and transitions should be labelled with there appropriate state or transition. * These are important for design as well as understanding

Answer 25

describe a ## Footnote **state diagram and its components**

Question 26

in regards to the context block what is the: ## Footnote **Information about the process’s memory space**

Answer 26

in regards to the context block: this is information about what memory the process has allocated to it. information recorded would include the address range of the memory

Question 27

the steps include: 1. Process A running 2. Process A makes SEM\_WAIT system call 3. Semaphore count is 0 4. Process A placed in waiting state 5. Process B makes SEM\_SIGNAL system call 6. Semaphore count is incremented by 1 7. Process A state changed from waiting to ready 8. Semaphore count set to 0 9. Process A is rescheduled and runs its critical code 10. Process A finishes and makes SEM\_SIGNAL system call 11. Semaphore count is incremented by 1 If step 3 had a count of 1 then execution would jump to step 8

Answer 27

describe in 11 steps what happens when: 1. process B is running critical code on a resource 2. process A tries to access the same resource and a mutex is in use by both processes

Question 28

this works in the following steps: 1. the opearating system has control of the CPU 2. it will set a hardware timer 3. a new process is scheduled to use the CPU 4. the hardware timer sends a hardware interrupt signal to the processor 5. the processor switches to kernel mode and the interrupt service routine for the timer is executed on the CPU 6. since the interrupt service routine for the timer is part of the operating system it can call on the scheduler if any process in a running state has quantum that has expired

Answer 28

describe how **pre-emptive multitasking** works in 6 steps

Question 29

two purposes of this are: 1. **for use by the scheduler** - so that it can determine which process should be executed next 2. **for use during a context switch** - so that a state of a process can be restored or saved (such as the state of registers or the memory a process has allocated to it)

Answer 29

name 2 purposes for the ## Footnote **context block**

Question 30

name 1 benefit and disadvantage of ## Footnote **pre-emptive multitasking**

Answer 30

a benefit of this is: If code is buggy or malicious then the operating system will always be able to regain control of the CPU a disadvantage of this is: that a process could be interrupted at a critical point in its execution and data could be left inconsistent. this is especially true of concurrent processes that are sharing data (race condition)

Question 31

describe in 8 steps the **stack pointer** during an interrupt **with** a context switch

Answer 31

this **with** a context switch 1. an interrupt occurs 2. process A has its instruction pointer placed on top the stack pointer 3. the routine is placed inside the instruction pointer and executed 4. the routine detects that process A has expired quantum and the OS scheduler is called 5. the operating system must make a context switch from process A to process B 6. the stack pointer register has its contents changed, now the stack pointer holds process B stack pointer 7. the routine finishes execution and calls the RETURN instruction 8. process B instruction pointer is pulled of the stack pointer and into the instruction pointer and execution commences where it was paused

Question 32

this is a phenomenon that occurs when two or more concurrent threads are sharing a resource its a problem that lies where they both are carrying out tasks on that resource but neither has finished its task fully the outcome is that depending on the order of who fnishes arrying out the task on that resource determines the final value of that resource the final value of the resource will most likely be incorrect

Answer 32

describe what the **race condition** is

Question 33

if a process is in this state then the operating system knows that it is waiting to be rescheduled to use the CPU

Answer 33

if a **process is in a ready state** what does the operating system know about the process

Question 34

outline the state of a process and any transitions for the following sceneraio **a process makes an I/O system call where it will be blocked and then rescheduled**

Answer 34

for this scenerio the steps are as follows: 1. the process A is in a running state 2. the process A makes an I/O system call to the operating system 3. processor mode switches to kernel mode 4. the operating system foresees the I/O operation taking some time 5. the process A is placed in a blocked state 6. a context switch occurs and another process is given control of the processor 7. processor mode switches to user mode 8. a hardware interrupt is received by the processor from the I/O device telling the processor that it is ready 9. processor mode switches to kernel mode 10. context switch occurs and the interrupt service routine for the I/O device is executed 11. the process A is unblocked and placed into a ready state until it is rescheduled

Question 35

what is a ## Footnote **semaphore**

Answer 35

this is a service that is built into the operating system. this enables the control and access of a shared resource so that threads sharing such a resource can avoid errors such as a race condition

Question 36

what is ## Footnote **context switching**

Answer 36

this is the act of the currently running process being replaced by another process _note_ since all data within the CPU registers will be replaced it is said that the CPU is now executing in a different "context"

Question 37

these signals are sent to the processor by an I/O device. telling the processor that it is ready to either send or recieve data

Answer 37

when are **hardware interrupt signals** sent to the processor

Question 38

name 2 uses of the ## Footnote **semaphore**

Answer 38

two use cases of this are: 1. **mutual exclusion (mutex)** - this can be used by a thread to tell other threads that it is currently executing a critical region of code and that the resource must not be acceessed this avoids scenerios such as a race condition 2. **buffer control** - this can also be used as a counter for example how many units are available in a buffer

Question 39

the operating system stores these inside of main memory and they are updated for each process whenever it makes a transition from a running state to any non running state

Answer 39

where does the ## Footnote **operating sytem store each context block for a process and when does each get updated**

Question 40

this **without** a context switch 1. an interrupt occurs 2. process A instruction pointer is placed on top the stack pointer 3. the routine is placed inside the instruction pointer and is executed 4. when the routine is finished it will call the RETURN instruction 5. process A instruction pointer is pulled of the stack pointer and execution commences where it was paused

Answer 40

describe in 5 steps the **stack pointer** during an interrupt **without** a context switch

Question 41

methods that will accomplish this are: * Cooperative multitasking * Pre-emptive multitasking

Answer 41

what are 2 methods that ## Footnote **ensure that the operating system can always gain back control of a processor so that it can carry out its own tasks such as scheduling**

Question 42

where does the ## Footnote **operating sytem store each context block for a process and when does each get updated**

Answer 42

the operating system stores these inside of main memory and they are updated for each process whenever it makes a transition from a running state to any non running state

Question 43

this **with** a context switch 1. an interrupt occurs 2. process A has its instruction pointer placed on top the stack pointer 3. the routine is placed inside the instruction pointer and executed 4. the routine detects that process A has expired quantum and the OS scheduler is called 5. the operating system must make a context switch from process A to process B 6. the stack pointer register has its contents changed, now the stack pointer holds process B stack pointer 7. the routine finishes execution and calls the RETURN instruction 8. process B instruction pointer is pulled of the stack pointer and into the instruction pointer and execution commences where it was paused

Answer 43

describe in 8 steps the **stack pointer** during an interrupt **with** a context switch

Question 44

in regards to the context block: this is information about what memory the process has allocated to it. information recorded would include the address range of the memory

Answer 44

in regards to the context block what is the: ## Footnote **Information about the process’s memory space**

Question 45

in 4 points how can a ## Footnote **race condition be avoided**

Answer 45

this can be avoided by: * identifiying shared resources * identifying critical code using those shared resources * employing methods that allow a thread to execute its critical code without being distured (semaphore) * employ methods that allow a thread exclusive access to a shared resource while it is carrying out critical code (semaphore)

Question 46

in regards to the context block what is the: ## Footnote **process state**

Answer 46

in regards to the context block: this is a record of what state the process is in such as running, ready or blocked, it is used to decide how the process should be treated.

Question 47

what are 2 methods that ## Footnote **ensure that the operating system can always gain back control of a processor so that it can carry out its own tasks such as scheduling**

Answer 47

methods that will accomplish this are: * Cooperative multitasking * Pre-emptive multitasking

Question 48

in the context of a race condition this can ensure that: only one thread can be executing critical code on a shared resource at a time this guarantees that a race condition cannot occur

Answer 48

in the context of a race condition what can a ## Footnote **semaphore ensure**

Question 49

in regards to the context block: this is a record of any open files that a process may have open and that the operating system needs to keep track of

Answer 49

in regards to the context block what is the: ## Footnote **Information about files**

Question 50

in regards to the context block: this uniquely identifies a process

Answer 50

in regards to the context block what is the ## Footnote **Process Identifier (PID)**

Question 51

what are 3 problems that arise from **context switching** which must always be solved to enable multitasking

Answer 51

3 problems that arise from this are: * The first is to ensure that the OS itself is executing when a context switch is required * second is what information about a process must be saved so that it can be stopped and restarted * third is how to cleanly switch the processor from executing one process to executing another

Question 52

when are **hardware interrupt signals** sent to the processor

Answer 52

these signals are sent to the processor by an I/O device. telling the processor that it is ready to either send or recieve data

Question 53

this is a service that is built into the operating system. this enables the control and access of a shared resource so that threads sharing such a resource can avoid errors such as a race condition

Answer 53

what is a ## Footnote **semaphore**

Question 54

There are also system calls that allow the creation and destroying of different semaphores so that different semaphores can be used by different applications

Answer 54

execpt for the semaphore system calls SEM\_WAIT and SEM\_SIGNAL ## Footnote **what other system calls are there that are associatied with the semaphore**

Question 55

if a **process is in a blocked state** what does the operating system know about the process

Answer 55

if a process is in this state then the operating system knows that it is waiting on an I/O operation to complete

Question 56

name 2 ## Footnote **disadvantages of Cooperative multitasking**

Answer 56

disadvantages of this include: 1. **process hogging** - programmers could choose to hold on to the CPU for longer than they should so that they can gain extra performance at the benefit of other processes 2. **buggy programs** - A buggy application that never reaches a point where control would be yielded back to the operating system such as being stuck in a loop. would potentially mean that the system would crash since the operating system cannot regain control of the CPU and make any recoveries

Question 57

in regards to the context block: this is the contents of the CPU registers that was stored for the process when a context switch occured saved data would include: * stack pointer register * instruction pointer register * status register

Answer 57

in regards to the context block what is the: ## Footnote **saved register contents**

Question 58

give an example of a race condition where the shared resource is a bank balance and thread A is depositing 100 and thread B is withdrawing 100

Answer 58

an example of this would be: 1. thread A reads the bank balance at 500 2. a context switch occurs 3. thread B raeds the banks balance at 500 4. a context switch occurs 5. thread A adds 100 to the 500 and updates the balance to 600 6. a context switch occurs 7. thread B subtracts 100 from the 500 and updates the balance to 400 the final balance is 400 when it should still be 500. this occured because thread A could not finish its critical code before the context switch

Question 59

this can be avoided by: * identifiying shared resources * identifying critical code using those shared resources * employing methods that allow a thread to execute its critical code without being distured (semaphore) * employ methods that allow a thread exclusive access to a shared resource while it is carrying out critical code (semaphore)

Answer 59

in 4 points how can a ## Footnote **race condition be avoided**

Question 60

describe what **Cooperative multitasking** is

Answer 60

this is a solution that enables multitasking it involves all application programs to be written in such a way that they will periodically yield control of the CPU back to the operating system so that it can be rescheduled and other processes can use the CPU.

Question 61

how many processors are operating systems designed to run on

Answer 61

these are designed to run on one processor. they are software just like any application program and must share control of the resources available. the crucial difference is that this has the responsibility to orchestrate the use of resources correctly and efficiently

Question 62

in regards to the context block what is the: ## Footnote **Information about files**

Answer 62

in regards to the context block: this is a record of any open files that a process may have open and that the operating system needs to keep track of

Question 63

when can **context switching** occur

Answer 63

this can only occur after a system call has been made or a hardware interrupt signal has been received. in either case the processor will switch to kernel mode and the OS will gain control of the CPU, allowing it a chance to check the quantum of any running processes and run its scheduler if any quantum has expired

Question 64

in regards to the context block what is the ## Footnote **Process Identifier (PID)**

Answer 64

in regards to the context block: this uniquely identifies a process

Question 65

3 problems that arise from this are: * The first is to ensure that the OS itself is executing when a context switch is required * second is what information about a process must be saved so that it can be stopped and restarted * third is how to cleanly switch the processor from executing one process to executing another

Answer 65

what are 3 problems that arise from **context switching** which must always be solved to enable multitasking

Question 66

what is a ## Footnote **critical region** **of code**

Answer 66

this is a place in the source code that acts on a shared resource. If these actions are not carried out correctly and in full then the value of the shared resource could become false

Question 67

disadvantages of this include: 1. **process hogging** - programmers could choose to hold on to the CPU for longer than they should so that they can gain extra performance at the benefit of other processes 2. **buggy programs** - A buggy application that never reaches a point where control would be yielded back to the operating system such as being stuck in a loop. would potentially mean that the system would crash since the operating system cannot regain control of the CPU and make any recoveries

Answer 67

name 2 ## Footnote **disadvantages of Cooperative multitasking**

Question 68

what is the main feature of the processor that the operating system takes advantage of to achieve ## Footnote **pre-emptive multitasking**

Answer 68

the main feature of the processor that the operating system takes advantage of to achieve this is: the fact that a hardware interrupt will pause the execution of any process then switch to kernel mode and hand control of the CPU to the operating system the hardware interrupt in question is received from a hardware timer

Question 69

describe the semaphore system call **SEM\_WAIT**

Answer 69

this system call essentially asks the OS if it can enter a critical region of code. If the semaphore count is 0 then the process is placed into a waiting state. If the semaphore count is not 0 then the semaphore count is set to 0 and the process enters a critical region of code

Question 70

the steps that describe this are: 1. a buffer is empty and has a size of 10 (the semaphore value is 10) 2. a process wishes to deposit an item so makes a SEM\_WAIT system call 3. since the semaphore is not 0 the process can deposit the item in the buffer 4. the semaphore value is now decremented to 9 5. steps 2 - 5 will reapeat untill the semaphore reaches 0 6. a process wishes to deposit an item so makes a SEM\_WAIT system call 7. the semaphor count is 0 and so the process is put into a waiting state until another process removes an item from the buffer and makes the system call SEM\_SIGNAL so that the semaphore can be incremented by 1 _note_ this example does not take into account an event such as two threads colliding at a buffer address (race condition) for this to be mitigated a seperate semaphore would be needed for mutual exclusion (mutex)

Answer 70

in 7 steps describe how ## Footnote **buffer control is possible by using a semaphore**

Question 71

This is a snapshot or a record of every process and its state that must be known about by the operating system

Answer 71

what is the ## Footnote **context block**

Question 72

while the processor is in this mode the opearting system can access: * all instructions including privileged instructions * all data structures including the context block

Answer 72

while the processor is in kernel mode what 2 items does the operating system have access to

Question 73

name 2 purposes for the ## Footnote **context block**

Answer 73

two purposes of this are: 1. **for use by the scheduler** - so that it can determine which process should be executed next 2. **for use during a context switch** - so that a state of a process can be restored or saved (such as the state of registers or the memory a process has allocated to it)

Question 74

this is a solution that enables multitasking it involves all application programs to be written in such a way that they will periodically yield control of the CPU back to the operating system so that it can be rescheduled and other processes can use the CPU.

Answer 74

describe what **Cooperative multitasking** is

Question 75

in two summary points when do Concurrency problems such as the race condition occur

Answer 75

in summary these problems occure when: 1. There is a shared resource 2. There are **critical regions** of code acting on those shared resources

Question 76

what is the ## Footnote **context block**

Answer 76

This is a snapshot or a record of every process and its state that must be known about by the operating system

Question 77

these are designed to run on one processor. they are software just like any application program and must share control of the resources available. the crucial difference is that this has the responsibility to orchestrate the use of resources correctly and efficiently

Answer 77

how many processors are operating systems designed to run on

Question 78

describe how **pre-emptive multitasking** works in 6 steps

Answer 78

this works in the following steps: 1. the opearating system has control of the CPU 2. it will set a hardware timer 3. a new process is scheduled to use the CPU 4. the hardware timer sends a hardware interrupt signal to the processor 5. the processor switches to kernel mode and the interrupt service routine for the timer is executed on the CPU 6. since the interrupt service routine for the timer is part of the operating system it can call on the scheduler if any process in a running state has quantum that has expired

Question 79

an example of this would be: 1. thread A reads the bank balance at 500 2. a context switch occurs 3. thread B raeds the banks balance at 500 4. a context switch occurs 5. thread A adds 100 to the 500 and updates the balance to 600 6. a context switch occurs 7. thread B subtracts 100 from the 500 and updates the balance to 400 the final balance is 400 when it should still be 500. this occured because thread A could not finish its critical code before the context switch

Answer 79

give an example of a race condition where the shared resource is a bank balance and thread A is depositing 100 and thread B is withdrawing 100

Question 80

if a **process is in a running state** what does the operating system know about the process

Answer 80

with this type of state the operating system knows that the process is currently using the CPU

Question 81

in regards to the context block what is the: ## Footnote **saved register contents**

Answer 81

in regards to the context block: this is the contents of the CPU registers that was stored for the process when a context switch occured saved data would include: * stack pointer register * instruction pointer register * status register

Question 82

what is a ## Footnote **mutex**

Answer 82

this is the term given to a semaphore when it is used for the purpose of mutual exclusion _note_ semaphores have many use cases within the operating system

Question 83

for this scenerio the steps are as follows: 1. the process A is in a running state 2. the process A makes an I/O system call to the operating system 3. processor mode switches to kernel mode 4. the operating system foresees the I/O operation taking some time 5. the process A is placed in a blocked state 6. a context switch occurs and another process is given control of the processor 7. processor mode switches to user mode 8. a hardware interrupt is received by the processor from the I/O device telling the processor that it is ready 9. processor mode switches to kernel mode 10. context switch occurs and the interrupt service routine for the I/O device is executed 11. the process A is unblocked and placed into a ready state until it is rescheduled

Answer 83

outline the state of a process and any transitions for the following sceneraio **a process makes an I/O system call where it will be blocked and then rescheduled**

Question 84

this system call essentially tells the operating system that it is finished executing its critical region of code and the semaphore count can be incremented so that another process may enter its critical region of code.

Answer 84

describe the semaphore system call SEM\_SIGNAL

Question 85

in this case it can guarantee mutual exclusion that is that only one process can be executing a critical region of code at the same time. this in turn can mitigate a race problem

Answer 85

when two values are being used for a semaphore ## Footnote **what can the semaphore guarantee**

Question 86

a benefit of this is: If code is buggy or malicious then the operating system will always be able to regain control of the CPU a disadvantage of this is: that a process could be interrupted at a critical point in its execution and data could be left inconsistent. this is especially true of concurrent processes that are sharing data (race condition)

Answer 86

name 1 benefit and disadvantage of ## Footnote **pre-emptive multitasking**

Question 87

describe in 11 steps what happens when: 1. process B is running critical code on a resource 2. process A tries to access the same resource and a mutex is in use by both processes

Answer 87

the steps include: 1. Process A running 2. Process A makes SEM\_WAIT system call 3. Semaphore count is 0 4. Process A placed in waiting state 5. Process B makes SEM\_SIGNAL system call 6. Semaphore count is incremented by 1 7. Process A state changed from waiting to ready 8. Semaphore count set to 0 9. Process A is rescheduled and runs its critical code 10. Process A finishes and makes SEM\_SIGNAL system call 11. Semaphore count is incremented by 1 If step 3 had a count of 1 then execution would jump to step 8

Question 88

in the context of a race condition what can a ## Footnote **semaphore ensure**

Answer 88

in the context of a race condition this can ensure that: only one thread can be executing critical code on a shared resource at a time this guarantees that a race condition cannot occur

Question 89

describe what **pre-emptive multitasking** is

Answer 89

this is a solution to multitasking within this solution the operating system has control over when it will use the CPU. it does this by setting a hardware timer that will interrupt any process

Question 90

describe a ## Footnote **state diagram and its components**

Answer 90

this is a diagram that expresses the transition and states of a process. * ovals show the state of a process * arrows show the direction and transition of a process from one state to another _notes_ * All states and transitions should be labelled with there appropriate state or transition. * These are important for design as well as understanding

Question 91

describe what the **race condition** is

Answer 91

this is a phenomenon that occurs when two or more concurrent threads are sharing a resource its a problem that lies where they both are carrying out tasks on that resource but neither has finished its task fully the outcome is that depending on the order of who fnishes arrying out the task on that resource determines the final value of that resource the final value of the resource will most likely be incorrect

Question 92

when two values are being used for a semaphore ## Footnote **what can the semaphore guarantee**

Answer 92

in this case it can guarantee mutual exclusion that is that only one process can be executing a critical region of code at the same time. this in turn can mitigate a race problem