Chapter 4

Question 1

Distinguish between hard real-time OS and soft real-time OS. Define jitter.

Answer 1

In hard real-time late completion of a task is inacceptable (i.e. nuclear reactor).

In soft real-time occasional late completion of a task is acceptable as long as the application generally meets the deadline. (i.e. office printer)

Variability of the time to complete an application task.

Question 2

Distinguish between server, desktop and general purpose operating systems.

Answer 2

Maximizes throughput (efficiency) and focuses on recoverability to failures and scalability.

Rapid startup, rapid user response times and energy efficiency.

General purpose aims to combine both to adapt to different applications.

Question 3

Explain round robin process scheduling algorithm.

Answer 3

All processes are advanced within a period of time. Each process runs for a short period (time quantum) before yielding CPU to the next process. Some time wasted by switching.

Question 4

What is the period of time each process runs for called?

Answer 4

Time Quantum

Question 5

What are the three different types of process?

Answer 5

Interactive: Requires quick response time.
Normal: Performs business-related computations and needs to maximize the usage of computer resources.
Housekeeping: Usually runs when there is nothing else to do.

Question 6

What is the disadvantage of basic round robin scheduling?

Answer 6

It doesn’t take into account the different kinds of task at all.

Question 7

Explain preemptive scheduling with priority queues (naive).

Answer 7

Priority assigned to each process. When picking next process:
1. Always pick the runnable process with the highest priority.
2. If two or more runnable processes with equal priority are available schedule them using round robin.
Processes are arranged into queues based on their priorities, with different time quantum for each as appropriate.

Question 8

What is the problem with preemptive scheduling with priority queues (naive).

Answer 8

If some high-priority process is always runnable, lower priority processes never get a chance to run. This is called starvation.

Question 9

Explain MFQS.

Answer 9

Multilevel Feedback Queue Scheduling.

1. If P(A) > P(B). A runs.
2. If P(A) = P(B), A & B run in RR using time quantum of given queue.
3. When job enters system, place at highest P.
4. Once a job uses up time allotment reduce its priority.
   5 After some time period S, move all the jobs in the system to the topmost queue.

Question 10

What is the downside of MFQS?

Answer 10

Has many parameters time quantum at different levels, how often to reset priorities, how to rebalance priorities. These are all set by admin called “voodoo constants” and are complex to understand or alter in an attempt to improve efficiency.

Question 11

Explain Linux completely fair scheduling nice value.

Answer 11

nice values range from -20 to 19. The lower the nice value - the higher the priority. The higher the nice value - the lower the priority.

Question 12

Explain Linux completely fair scheduling virtual runtime.

Answer 12

Virtual runtime of a process is the total amount of CPU time it has received so far.

Question 13

Explain Linux completely fair scheduling.

Answer 13

When choosing the next process to run always choose the runnable process with the minimal virtual runtime. Stores virtual runtime in RB tree for efficiency (leftmost node minimum). When a new process is started, its inserted at this leftmost node to give it the highest priority. Nice value of process affects how much virtual runtime it accumulates. So a lower nice value leads to less accumulation and increases priority.

Question 14

Define BIOS.

Answer 14

Basic Input Output System.
Initial program stored in ROM
Loads OS kernel
Provides run time services for OS.
Performs POST (Power-on self test).
Performs required hardware initialization.

Question 15

What does UART stand for?

Answer 15

Universal Asynchronous Receiver/Transmitter

Question 16

Where is OS kernel (supervisor) in QEMU system?

Answer 16

Executable file loaded into RAM

Question 17

What is booting?

Answer 17

Refers to computer startup and involves:
1. Loading OS kernel into memory
2. Initializing computer hardware adapters and peripherals
3. Starting background application programs
4. Starting interactive shell

Question 18

What is the typical boot loading stages in linux?

Answer 18

Bios Bootloader Program runs in ROM -> loads from data storage ->
Operating System Bootloader runs in RAM -> loads from data storage
-> Operating System Kernel runs in RAM -> launches from file system -> init file runs -> starts all other programs (e.g. shell and daemon processes)

Question 19

What is the access control matrix?

Answer 19

Rows represent Users, columns objects, and data the operations the corresponding user has permission to perform on the corresponding object.

Question 20

What do access control mechanisms do?

Answer 20

Only allow users with the appropriate permissions to perform appropriate actions.

Question 21

How are user groups set? Why key information is used by access control mechanisms that is stored in User account?

Answer 21

On a multi-user computer system, users are split into several groups based on their job function. Data sharing between groups should be limited.

User identifier (UID)
Group identifier (GID)

Question 22

What is superuser?

Answer 22

1. Special user account called root
2. Has privileges above all other accounts
3. Can change ownership and permissions
4. Has full access to all files
5. Can kill any process

Question 23

What is sudo group?

Answer 23

The superuser access is too dangerous for interactive login, but users part of the sudo group can access its privileges.

Question 24

How are file permissions specified?

Answer 24

For file owner (match UID), for group of owner (match GID) and for all other users

drwxrwxrw

first letter is file type (d=directory, -=ordinary file, l=link, b=block device etc.)
r = permission to read contents
x = permission to execute/search
w = permission to modify
- = No permission

Question 25

What is a semaphore?

Answer 25

An unsigned integer variable which is initialized to some value >= 0. Binary semaphore is similar to a lock. wait operation attempts to decrement the value of the semaphore variable by 1. If new value would be negative go to sleep, otherwise continue execution. signal operation attempts to increment the value of the semaphore by 1. If there are processes sleeping on the semaphore it wakes them up.

Question 26

Suppose we have three threads A, B and C and three critical sections for each how can we go through them in order.

Answer 26

Sa=1, Sb = 0 Sc = 0 wait(Sa) critical section A signal(Sb) wait(Sb) critical section B signal(Sc) wait(Sc) critical section C signal(Sa)

Question 27

How can we use semaphores to keep track of used GPU's?

Answer 27

Initialize a semaphore to the number of GPUS free, then call function wait() and mark that GPU as used. This will decrement semaphore. Have another semaphore to make sure the part of the code that changes GPU count can only be accessed by one GPU at a time. Have another function to release semaphore.

Question 28

How is data allocated on the stack?

Answer 28

By decrementing stack pointer by number of bytes.

Question 29

In which direction does stack grow?

Answer 29

Downward

Question 30

In a typical C program memory layout, which section is used to store uninitialized global and static variables?

Answer 30

.bss

Question 31

In a single multitasking CPU hart, how can interrupts be used to prevent race conditions when accessing shared resources like an MMIO display, as discussed in the context of critical sections?

Answer 31

By disabling interrupts before entering a critical section and re-enabling them upon exit.

Question 32

What is the primary function of a Memory Management Unit (MMU)?

Answer 32

To translate virtual addresses generated by the CPU into physical memory addresses.

Question 33

In Unix, what is a ”file descriptor”?

Answer 33

A small, non-negative integer that the kernel uses to identify an open file for a process.

Question 34

In a Unix shell, what is the typical purpose of input/output redirection using > (greater-than sign)?

Answer 34

To redirect the standard output of a command to a file, overwriting the file if it exists.

Question 35

Which of the following best describes the sbrk() system call in traditional Unix systems?

Answer 35

To change the size of the data segment (heap) of the calling process.

Question 36

In the cut command, when using field-based extraction (e.g., cut -d’ ’ -f FIELDS file.txt), what does the field specifier -f2- typically mean?

Answer 36

To select all fields from the second field to the end of the line.

Question 37

For many system information utilities (like free or tools reading /proc/cpuinfo) to function correctly in a minimal Linux environment (e.g., booted with an initramfs), what filesystem typically needs to be mounted?

Answer 37

/proc (process information pseudo-filesystem)

Question 38

What does the uname -a command typically display in a Linux environment?

Answer 38

All system information including kernel name, hostname, kernel release, kernel version, machine hardware name, and operating system.

Question 39

Which of these is a common strategy to avoid deadlocks when multiple locks are involved?

Answer 39

Acquiring locks in a predefined global order and releasing them in the reverse order.

Question 40

In the context of QEMU, if the -bios none option is used, what typically happens at the start of emulation?

Answer 40

QEMU directly loads the specified kernel into RAM, and the emulated ROM contains minimal code to jump to the kernel’s entry point.