Lecture 2

Question 1

Steps in the transition from user to kernel mode

Answer 1

• User code issues a system call
• Mode is changed to kernel mode
• OS checks that everything is in order and legal
• OS executes system call
• Mode is set back to user mode

Question 2

True or false: a single system call may contain multiple privileged instructions

Answer 2

True

Question 3

System calls enable users to…

Answer 3

…Access OS services

Question 4

System calls are typically written…

Answer 4

…in a high-level programming language (like C/C++)

Question 5

Each system call has a…

Answer 5

…System call number

Question 6

System calls are accessed via…

Answer 6

…an API (Application Program Interface)

Question 7

What does an API do

Answer 7

• Hides the details of implementation of system calls

- Offers portability across the OS

Question 8

Common APIs

Answer 8

• POSIX (UNIX, Linux, Mac OS X)
• Java API (Java Virtual Machine)
• Win32 (Windows)

Question 9

Programming languages can be seen as APIs when they have…

Answer 9

…file I/O operations

Question 10

What does the command $ strace -c [command] do?

Answer 10

Displays summary info on system calls invoked during

the execution of the command

Question 11

Numerous system calls are typically invoked for…

Answer 11

…even simple tasks

Question 12

Types of system calls

Answer 12

• Process Control
• File Manipulation
• Device Manipulation
• Information Maintenance
• Communication
• Protection

Question 13

Steps in using a timer for protection

Answer 13

• Before giving user control of a program, OS sets a
  timer (usually in the millisecond range)
• After period expires, timer chip issues a system
  interrupt, giving the OS control again
• OS decides whether to kill the process or let it
  continue

Question 14

Why is a user holding resources to long a problem?

Answer 14

Malicious user might be exploiting something, or the program may have just bugged out

Question 15

Types of OS structure

Answer 15

• Monolithic (everything the OS does goes in the
  kernel)
• Layered (Hardware is bottom layer, user interface is
  top layer, with multiple layers in between)
• Microkernel (Kernel only does what it absolutely has to)
• Modular
• Hybrid

Question 16

Pros and Cons of Monolithic OS

Answer 16

• Pro: efficient in terms of space and speed
• Con: Difficult to implement + maintain (too much in
  one layer)

Question 17

Pros and Cons of Layered OS

Answer 17

• Pro: Easier to develop, debug, and update
• Con: Less efficient, and it’s hard to decide what goes
  on which layer

Question 18

In a microkernel structure, modules communicate using…

Answer 18

…Message passing through kernel

Question 19

Pros and Cons of Microkernel OS

Answer 19

• Pro: Easier to extend microkernel/port to new devices,
  more reliable and secure
• Con: More performance overhead

Question 20

In a modular structure, core components go in the…

Answer 20

…Kernel module

Question 21

In a modular OS, how are new modules loaded?

Answer 21

They are linked to the kernel as needed

Question 22

Modules talk to each other using…

Answer 22

…well-defined interfaces

Question 23

Modular Structure Advantages

Answer 23

• Easy to maintain, update, and debug
• Similar to layers but more flexible
• Similar to microkernel but more efficient

Question 24

OS must be available to ____ to start it

Answer 24

Hardware

Question 25

On powering up, the instruction register is loaded with…

Answer 25

…a predefined memory location (usually the location of the Bootstrap loader in ROM)

Question 26

Bootstrap loader routine:

Answer 26

• Performs diagnostic tests
• Move boot block from fixed location to memory
• Boot block loads the rest of the loader from the disk
• Boostrap loader loads kernel