Multiple CHoice QUestions

Question 80

Some filesystems require ______________ tools to restore the performance on mechanical drives, which have sections of the filesystem become non-contiguous.

Save

Question 81 (1 point)

Question 81 Unsaved

Recording all the steps that are about to be taken prior to taking the steps, called __________, solves the problem of a computer crashing or losing power, leading to a file system entering an inconsistent state.

Save

Question 82 (1 point)

Question 82 Unsaved

______ are data structures on Unix systems that contain file metadata and pointers to the actual data, but not the actual file name. _______ are data structures that map filenames to inode numbers.

Blank # 1

Blank # 2

Save

Question 83 (1 point)

Question 83 Unsaved

A file system is made available to users and programs on a computer systems by the process of ________ a drive.

Save

Question 84 (1 point)

Question 84 Unsaved

CPUs switch process quickly, enabling all of them to make forward progress, while providing the illusion that all processes are running at the same time. This is called _______________.

Save

Question 85 (1 point)

Question 85 Unsaved

To access all memory on a system, the OS must be in ______ mode.

Save

Question 86 (1 point)

Question 86 Unsaved

An ________ happens when currently executing code is stopped so the operating system can handle and incoming event. The can be involuntary, for example from I/O, or voluntary, from system calls or exceptions.

Save

Question 87 (1 point)

Question 87 Unsaved

Match the following functions of a kernel, with the type of function it falls under.

12

prevent a single running program from taking over the system

12

inter process communication

12

enforce access privileges

12

mechanism for programs to access hardware

1.

abstraction

2.

arbitration

Save

Question 88 (1 point)

Question 88 Unsaved

Match the term with the appropriate definition

12

software methods that enable operations to be carried out

12

software methods that enforce permissions, access rules, or other limits against applications

1.

mechanism

2.

policy

Save

Question 89 (1 point)

Question 89 Unsaved

Match the term with the appropriate definition

12

software methods that enforce permissions, access rules, or other limits against applications

12

software methods that enable operations to be carried out

1.

mechanism

2.

policy

Save

Question 90 (3 points)

Question 90 Unsaved

Match the kernel type to its correct properties

12

rest of OS is divided into collection of servers running in userspace, generally with lower priveleges

12

different functions of the OS are divided into subsystems

12

entire OS is placed in kernel space

12

all OS code runs in privileged mode (ring 0 on x86)

12

higher performance at the expense of poorer modularity and separation of components

12

bare minimum of code runs in kernel space (basic address, IPC, and scheduling)

12

excellent modularity, but performance suffers

1.

monolithic kernel

2.

microkernel

Save

Question 91 (1 point)

Question 91 Unsaved

Most kernels today are purely monolithic.

True

False

Save

Question 92 (1 point)

Question 92 Unsaved

Most kernels today are purely microkernels.

True

False

Save

Question 93 (1 point)

Question 93 Unsaved

Most kernels today are a hybrid of monolithic kernels and microkernels.

True

False

Save

Question 94 (1 point)

Question 94 Unsaved

Memory allocation algorithms always allocate exactly the amount of memory requested.

True

False

Save

Question 95 (1 point)

Question 95 Unsaved

External memory fragmentation refers to when many allocations do not use an entire block of allocated memory, resulting in wasted space.

True

False

Save

Question 96 (1 point)

Question 96 Unsaved

The best fit algorithm for memory allocation, which finds and allocates the smallest chunk of free space that is available to accommodate the request, is efficient.

True

False

Save

Question 97 (1 point)

Question 97 Unsaved

Allocating memory by finding the first fit, that is find and allocating the first chunk of memory that is large enough to satisfy the request, results in small fragments of free space at the start of the free list, degrading performance over time.

True

False

Save

Question 98 (1 point)

Question 98 Unsaved

Binary tree based memory allocation methods are more efficient than linked list based methods.

True

False

Save

Question 99 (1 point)

Question 99 Unsaved

Power of two methods of memory allocation increase external fragmentation.

True

False

Save

Question 100 (1 point)

Question 100 Unsaved

A larger lower limit on the block size in a power of two memory allocation algorithm produces less internal fragmentation.

True

False

Save

Question 101 (1 point)

Question 101 Unsaved

Coalescence, or combining neighboring free blocks when a block is freed, reduces internal fragmentation.

True

False

Save

Question 102 (1 point)

Question 102 Unsaved

______ provides a mechanism to facilitate sharing memory among multiple user space processes at the same time.

Save

Question 103 (1 point)

Question 103 Unsaved

A page refers to block of _______ (blank 1) memory and a frame refers to a block of ________ (blank 2) memory.

Blank # 1

Blank # 2

Save

Question 104 (1 point)

Question 104 Unsaved

The memory managment unit of a CPU translates a ______ memory address into a _______ memory address.

Blank # 1

Blank # 2

Save

Question 105 (1 point)

Question 105 Unsaved

The ____ table maps page numbers to frame numbers and is stored in physical memory (in the kernel space)

Save

Question 106 (1 point)

Question 106 Unsaved

Since each page translation requires a table lookup, two memory accesses are required for each data access and for each instruction, which is inefficient. To solve this problem, the ___ (three letter acronym) is an associative memory which provides a shortcut to speed up translation.

Save

Question 107 (1 point)

Question 107 Unsaved

When an address is found in the TLB, it is known as a ___. (one word)

Save

Question 108 (1 point)

Question 108 Unsaved

Hierarchical page tables divide page tables into pages. This results in a logical memory address containing how many components?

5

4

3

2

Save

Question 109 (1 point)

Question 109 Unsaved

When a process accesses logical memory, the CPU translates the logical address into a physical address using its ___ (3 letter acronym).

Save

Question 110 (1 point)

Question 110 Unsaved

Order the steps that occur when a process accesses a logical memory address.

123

The appropriate physical memory location is accessed.

123

The MMU consults the page table for the frame number corresponding to the given page number.

123

The MMU constructs the physical address by using the frame number and the offset given by the logical address (left unchanged).

Save

Question 111 (1 point)

Question 111 Unsaved

When a process accesses memory, it sees the frame number of the memory location being accessed.

True

False

Save

Question 112 (1 point)

Question 112 Unsaved

What are the parts of a logical memory address for a system with hierarchical address translation?

Question 112 options:

memory address

outer page table

inner page table offset

frame offset

inner page number

physical address

Save

Question 113 (1 point)

Question 113 Unsaved

Hashed address translation uses a hash function to directly translate a logical address into a physical address.

True

False

Save

Question 114 (1 point)

Question 114 Unsaved

Hashed address translation never results in collisions of a page mapping to multiple frames.

True

False

Save

Question 115 (1 point)

Question 115 Unsaved

Inverted page tables store one entry per frame, which maps to to corresponding process ID and page number. This makes inverted page tables are inefficient for storage, but efficient for speed.

True

False

Save

Question 116 (1 point)

Question 116 Unsaved

Access restrictions apply to each region of process memory. For example, the text segment is read-only.

True

False

Save

Question 117 (1 point)

Question 117 Unsaved

The segment table has permission bits to each region of memory. A valid/invalid bit is used to prevent a process from writing to a segment belonging to a different process.

True

False

Save

Question 118 (1 point)

Question 118 Unsaved

When a process attempts to access an invalid page, a ___________ fault results.

Save

Question 119 (1 point)

Question 119 Unsaved

Programs such as web browsers can share read-only pages in memory (eg. multiple tabs can share the text segment of memory).

True

False

Save

Question 120 (1 point)

Question 120 Unsaved

If two processes are sharing a region of memory, they will have the same page number and frame number for that region of memory.

True

False

Save

Question 121 (1 point)

Question 121 Unsaved

If two processes are sharing a region of memory, they will have the invalid bit set.

True

False

Save

Question 122 (1 point)

Question 122 Unsaved

Virtual memory allows logical memory space to be larger than physical memory space.

True

False

Save

Question 123 (1 point)

Question 123 Unsaved

The process of moving pages between physical memory and a backing store, such as disk, is known as ______.

Save

Question 124 (1 point)

Question 124 Unsaved

Modern linux systems require a separate disk partition for swapping pages.

True

False

Save

Question 125 (1 point)

Question 125 Unsaved

Modern linux systems require the existence of a swap file.

True

False

Save

Question 126 (1 point)

Question 126 Unsaved

A ____ fault occurs when a process accesses a logical memory page that is not resident in physical memory.

Save

Question 127 (2 points)

Question 127 Unsaved

Order the following steps in page fault handling appropriately

1234

OS page fault handler swaps page in from disk and updates the page table accordingly.

1234

MMU checks TLB then page table, if necessary. If page is loaded, translate frame number and access memory. If page is not loaded, MMU raises an interrupt, causing CPU to switch context to the OS page fault handler.

1234

OS switches CPU context back to the requesting process and memory access proceeds normally.

1234

Process makes a memory request in logical address space

Save

Question 128 (1 point)

Question 128 Unsaved

Pages swapped to the backing store (such as the hard disk) are secure and cannot be recovered by forensic methods.

True

False

Save

Question 129 (1 point)

Question 129 Unsaved

When the fraction of memory access resulting in page faults becomes too high, system performance will decrease significantly. This is known as __________. (one word answer)

Save

Question 130 (1 point)

Question 130 Unsaved

One way to diminish thrashing is ___________, where some pages are loaded into memory before they are actually referenced. Though this increases memory use, it can increase performance.

Save

Question 131 (1 point)

Question 131 Unsaved

Unix processes are created by cloning another process using ____ system call.

Save

Question 132 (1 point)

Question 132 Unsaved

Cloning all memory when a process is forked would incur significant performance overhead. ____ __ _____ is a solution to have both the original and forked processes initially share memory pages. (3 words in the blank)

Save

Question 133 (1 point)

Question 133 Unsaved

Many processes perform frequent reads and writes to disk with small data sizes, which has a large performance overhead. Rather than waiting for this I/O for slow devices such as mechanical hard disks or SSDs, ______ ______ files allow for performance increases by mapping page-sized pieces of the file into memory. (2 words)

Save

Question 134 (1 point)

Question 134 Unsaved

Shared libraries facilitate sharing of certain routines to many different programs. _______ linking refers to the loader adding these shared library routines at runtime. (1 word)

Save

Question 135 (1 point)

Question 135 Unsaved

Shared libraries facilitate sharing of certain routines to many different programs. _______ linking refers to the linker adding these shared library routines at compile time. (1 word)

Save

Question 136 (1 point)

Question 136 Unsaved

.so files in Unix and .dll files on Windows are compiled shared libraries stored on disk.

True

False

Save

Question 137 (1 point)

Question 137 Unsaved

Page replacement is performed by the OS to remove pages from memory to allow for newly required data to be placed in memory.

True

False

Save

Question 138 (1 point)

Question 138 Unsaved

Under a local page replacement policy, a process can steal frames from another process.

True

False

Save

Question 139 (1 point)

Question 139 Unsaved

Under a global page replacement policy, process are allocated a limited number of frames and only frames belonging to a given process can be selected for replacement.

True

False

Save

Question 140 (1 point)

Question 140 Unsaved

Allocating more frames to a process will always produce fewer page faults.

True

False

Save

Question 141 (1 point)

Question 141 Unsaved

The minimum number of frames needed is called the _______ ___ of the process. (2 words)

Save

Question 142 (1 point)

Question 142 Unsaved

The least recently used page replacement algorithm works will in theory and in practice.

True

False

Save

Question 143 (1 point)

Question 143 Unsaved

Match the following process states with its description

12345

process is executing

12345

process is waiting to be assigned to a CPU core

12345

process is being created

12345

process has finished execution

12345

process is waiting for some event

1.

new

2.

ready

3.

running

4.

waiting

5.

terminated

Save

Question 144 (1 point)

Question 144 Unsaved

All processes are created from a parent process, except for _____.

Save

Question 145 (1 point)

Question 145 Unsaved

Child processes must wait until the parent terminates.

True

False

Save

Question 146 (1 point)

Question 146 Unsaved

When a child is spawned by the parent process, the parent can no longer terminate the child process.

True

False

Save

Question 147 (1 point)

Question 147 Unsaved

A _______ switch occurs when a system switches from running a process to running kernel code. This is a computationally expensive operation.

Save

Question 148 (1 point)

Question 148 Unsaved

A _______ switch, where the OS switches from running one process to another, requires two context switches into the kernel, first to save the state of the previous process and second to restore the state of a second process.

Save

Question 149 (1 point)

Question 149 Unsaved

Match the following terms with the correct defintion

12

amount of work done in a fixed time

12

time elapsed between start and end of a program

1.

response time

2.

throughput

Save

Question 150 (1 point)

Question 150 Unsaved

Match the following left and right sides of the equations

1234

1/execution time

1234

clock cycle time * number of instructions * average clock cycles per instruction

1234

power * time

1234

number of instructions * average clock cycles per instruction

1.

performance

2.

CPU clock cycles

3.

execution time

4.

energy

Save

Question 151 (1 point)

Question 151 Unsaved

For each of the following quantities, match it with the factor affecting performance

123

cycles per instruction

123

number of instructions

123

clock cycle time

1.

how fast is each transistor / manufacturing process

2.

quality of the compiler and instruction set architecture

3.

the nature of the instruction and the quality of the architecture implementation

Save

Question 152 (1 point)

Question 152 Unsaved

How does SPEC compute its performance benchmark over the 29 test applications?

arithmetic mean

geometric mean

weighted arithmetic mean

Save

Question 153 (1 point)

Question 153 Unsaved

How much faster is a system with a SPEC 800 rating versus one with a SPEC 400 rating?

2x

400x

log(800/400)

4x

Save

Question 154 (1 point)

Question 154 Unsaved

Performance improvements through an enhancement is limited by the fraction of time the enhancement comes into play is known as

's Law. Enter a single word.

Save

Question 155 (1 point)

Question 155 Unsaved

Match the following principles affecting performance

123

10% of the program accounts for 90% of the execution time

123

the same code/data will be used again

123

nearby data/code will be touched next

1.

90-10 rule

2.

temporal locality

3.

spatial locality

Save

Question 156 (1 point)

Question 156 Unsaved

System A is 2 times faster than system B

How many seconds would a program run on System B that takes 10 seconds on system A? Enter only an integer in the first blank.

What is the percentage execution time increase/decrease for compared to system A? Enter only an integer (no percent sign) in the second blank.

Blank # 1

Blank # 2

Save

Question 157 (1 point)

Question 157 Unsaved

Which of the following two systems is better?

A program is converted to 5 billion instructions by a compiler. The MIPS processor is implemented such that each instruction completes in an average of 2 cycles and the clock speed is 1GHz

processor is implemented such that each instruction completes in an average of 6 cycles and the clock speed is 2GHz

Save

Question 158 (1 point)

Which of the following two systems is better?

A program is converted to 2 billion instructions by a compiler. The MIPS processor is implemented such that each instruction completes in an average of 2 cycles and the clock speed is 1GHz

processor is implemented such that each instruction completes in an average of 6 cycles and the clock speed is 2GHz

Save

Question 159 (1 point)

Which of the following two systems is better?

Question 159 options:

A program is converted to 2 billion instructions by a compiler. The MIPS processor is implemented such that each instruction completes in an average of 2 cycles and the clock speed is 1GHz

processor is implemented such that each instruction completes in an average of 2 cycles and the clock speed is 2GHz

Save

Question 160 (1 point)

If a 2Ghz process graduates an instruction every 4th cycle, how many billion instructions are there in a program that runs for 10 seconds? (for an answer, of 99 billion instructions, just enter 99 in the blank)

Save

Question 161 (1 point)

If a 4Ghz process graduates an instruction every 4th cycle, how many billion instructions are there in a program that runs for 10 seconds? (for an answer, of 99 billion instructions, just enter 99 in the blank)

Save

Question 162 (1 point)

Question 162 Unsaved

If a 4Ghz process graduates an instruction every 2nd cycle, how many billion instructions are there in a program that runs for 10 seconds? (for an answer, of 99 billion instructions, just enter 99 in the blank)

Save

Question 163 (1 point)

Question 163 Unsaved

If a program runs for 20 seconds on a 3 Ghz processor, how many clock cycles did it run for? (for an answer, of 99 billion cycles, just enter 99 in the blank)

Save

Question 164 (1 point)

If a program runs for 20 seconds on a 2 Ghz processor, how many clock cycles did it run for? (for an answer, of 99 billion cycles, just enter 99 in the blank)

Save

Question 165 (1 point)

Question 165 Unsaved

If a program runs for 10 seconds on a 2 Ghz processor, how many clock cycles did it run for? (for an answer, of 99 billion cycles, just enter 99 in the blank)

Save

Question 166 (1 point)

Question 166 Unsaved

A web server spends 30% of time in the CPU and 70% of time doing I/O. What is the maximum execution time reduction? Enter only an integer (with no percent sign)

Save

Question 167 (1 point)

Question 167 Unsaved

A web server spends 30% of time in the CPU and 70% of time doing I/O. What is the reduction in execution time for a processor that is 10 times faster? Enter only an integer (with no percent sign)

Save

Question 168 (1 point)

Question 168 Unsaved

A web server spends 20% of time in the CPU and 80% of time doing I/O. What is the reduction in execution time for a processor that is 10 times faster? Enter only an integer (with no percent sign)

Save

Question 169 (1 point)

Question 169 Unsaved

A web server spends 60% of time in the CPU and 40% of time doing I/O. What is the reduction in execution time for a processor that is 10 times faster? Enter only an integer (with no percent sign)

Save

Question 170 (1 point)

Question 170 Unsaved

A web server spends 80% of time in the CPU and 20% of time doing I/O. What is the reduction in execution time for a processor that is 10 times faster? Enter only an integer (with no percent sign)

Save

Question 171 (1 point)

Question 171 Unsaved

Match the term with the appropriate definition and properties

12

OS periodically queries each device to see if new information is available

12

Increased CPU usage due to lack of events from device

12

Device sends a signal to the OS to request attention and the OS preempt the running process to handle the device request

12

More complex to implement since it requires hardware support

12

High latency due to device waiting for the CPU

12

Allows for responsive system

1.

polling

2.

interrupt

Save

Question 172 (1 point)

Question 172 Unsaved

Different levels of interrupts allow the CPU to prioritize interrupts which occur simultaneously.

True

False

Save

Question 173 (1 point)

Question 173 Unsaved

Match the following terms to the appropriate definition.

123

Cache must get the block from memory

123

Cache contains the block being searched for

123

A new block is loaded into the cache

1.

Cache hit

2.

Cache miss

3.

Cache replacement

Save

Question 174 (1 point)

Question 174 Unsaved

address/data bus

Consider the above diagram of the address/data bus for reading/writing from memory in a computer system. Order the following steps appropriately for reading data from memory to the CPU.

123

Memory responds and puts data on to the "DATA" line

123

CPU drives "READ" line high

123

CPU puts the address out on to "ADDR" lines

Save

Question 175 (1 point)

Question 175 Unsaved

address/data bus

Consider the above diagram of the address/data bus for reading/writing from memory in a computer system. Order the following steps appropriately for reading data from memory to the CPU.

1234

CPU puts the address out on to "ADDR" lines

1234

CPU drives "WRITE" line high

1234

CPU puts data on the "DATA" line

1234

Memory responds and stores the data

Save

Question 176 (1 point)

Question 176 Unsaved

The stack stores allocated variables, structures, and objects, rather than automatic variables.

True

False

Save

Question 177 (1 point)

Question 177 Unsaved

Match the following definitions

1234

cache waits - updates the cache, but the cache doesn't get copied into memory until the cache block is evicted

1234

cache contains the block getting updated and the cache immediately forwards update to main memory

1234

cache doesn't contain the block and will read and update the block into the cache

1234

cache will not load the block update the block in the cache, but will forward the write straight to memory

1.

write-through

2.

write-back

3.

write-allocate

4.

write-no-allocate

Save

Question 178 (1 point)

Question 178 Unsaved

For the write-back policy, in which a write updates the copy in cache, but is not written through to memory a _____ bit is needed to indicate whether a block has been changed.

Save

Question 179 (1 point)

Question 179 Unsaved

On a cache write-miss, the typical policy used is?

write-allocate

write-no-allocate

Save

Question 180 (1 point)

Question 180 Unsaved

In an associative memory, data is guaranteed to be returned for a given key.

True

False

Save

Question 181 (1 point)

Question 181 Unsaved

Given a block size of 64 bytes and 512 cache lines the cache size is how any kilobytes? (enter an integer)

Save

Question 182 (1 point)

Question 182 Unsaved

Define the following types of cache misses

123

cache is too small to contain the entire working set

123

plenty of room in the cache but 2 blocks in the working set cannot be in the cache together

123

nothing in the cache when program begins

1.

compulsory miss

2.

capacity miss

3.

conflict miss

Save

Question 183 (1 point)

Question 183 Unsaved

When every access to the cache results in a miss is known as _________. (one word)

Save

Question 184 (1 point)

Question 184 Unsaved

In associative memories, the keys must be a fixed size, as in traditional memories.

Question 184 options:

True

False

Save

Question 185 (1 point)

Question 185 Unsaved

The computer moves data back and forth from main memory and cache memory in units called _______. (one word)

Save

Question 186 (1 point)

Question 186 Unsaved

A cache is made up of _____, each of which stores a block along with a tag and possibly a valid/dirty bit.