Section A

Assume three (3) processes (Process 1, 2, and 3), which share access to two semaphores U and V, as follows:

In each process, all commands are sequentially executed. Processes start their execution, request access to their critical section controlled by the acquire() method, and leave their critical section with the release() method. The method print('x') prints to the standard output (e.g., display/monitor) the character 'x'. The 'goto L' command transfers the control of the process execution to the command attached to the label 'L:'. Note: It is assumed that all the processes run (are scheduled) as often as possible; i.e. the semaphores are the only constraints on process execution.

1. How many characters 'C' are going to be printed?

Exactly four (4) characters 'C' will be printed

Exactly three (3) characters 'C' will be printed

No more than two (2) characters 'C' will be printed

At most four (4) characters 'C' will be printed

2. How many characters 'D' are going to be printed?

No more than two (2) characters 'D' will be printed.

Exactly three (3) characters 'D' will be printed.

At most four (4) characters 'D' will be printed.

Exactly four (4) characters 'D' will be printed.

3. Which is the minimum number of characters 'A' that will be printed?

The minimum number of 'A' printings is zero (0).

The character 'A' will be printed at least two (2) times, thus, the minimum number is two (2).

The character 'A' will be printed exactly one (1) time, thus, the minimum number is one (1).

The character 'A' will be printed exactly three (3) times, thus, the minimum number is three (3).

4. Is the sequence of the characters: 'CABABDDCABCABD' possible to be printed (with exactly this order)?

This exact sequence is possible to be printed if and only if the semaphore U is initialized to four (4)

Yes, this exact sequence is possible to be printed.

This exact sequence is possible to be printed if and only if the semaphore U is initialized to fourteen (14), i.e., as the number of characters in this sequence.

No, this exact sequence will never be printed; it's impossible.

Now consider the following cooperating processes A and B that share the (common) variable X:

The initial value of X is 5 (X=5;) before any process execution. The commands within each process are executed sequentially.

5. How many different values can the variable X get after the execution of both processes? Select one of the following:

There are six (6) different values that X will get after the execution of both processes.

There is only one (1) value that X will get after the execution of both processes.

There are two (2) different values that X will get after the execution of both processes.

There are four (4) different values that X will get after the execution of both processes.

Let us modify both programs by adding a semaphore S, which is initialized to 1 (S=1;), and X is initially 5 (X=5;). That is:

6. How many different values can the variable X get after the execution of both processes? Select one of the following:

There is only one (1) value that X will get after the execution of both processes.

There are four (4) different values that X will get after the execution of both processes.

There are six (6) different values that X will get after the execution of both processes.

There are two (2) different values that X will get after the execution of both processes.

Let us modify again the initial programs by adopting a semaphore T, which is initialized to 0 (T=0;), and X is initially 5 (X=5;). That is:

7. How many different values can the variable X get after the execution of both processes? Select one of the following:

There are two (2) different values that X will get after the execution of both processes.

There is only one (1) value that X will get after the execution of both processes.

There are six (6) different values that X will get after the execution of both processes.

There are four (4) different values that X will get after the execution of both processes.

Assume there are processes P and Q with the semaphores s and t, both initialized to 1 (s=1; t=1). Which of the following sequences of the commands P1, ..., P8, Q1, ..., Q5, leads those processes to a deadlock state, if any?

8. Select one of the following:

The sequence P1, P2, P3, P4, Q1, Q2, P5 results in a set of deadlocked processes.

The sequence P1, P2, P3, P4, P5, P6, P7, Q1, P8 results in a set of deadlocked processes.

There will be no deadlock with any sequence of the commands.

The sequence Q1, Q2, Q3, Q4, P1, Q5, P2 results in a set of deadlocked processes.

Section B

Imagine a system with an 8-bit byte-addressable physical address space.

9. How many bytes of memory will there be?

256

10. For this system, consider using a virtual addressing scheme with single-level paging. If each page contains 16 bytes, how many pages will there be?

256

11. In the worst case, what happens to memory access latency in a virtual addressing environments with a single-level page table, with respect to physical addressing?

Access time quadruples

Access time is unbounded

Access time doubles

Access time halves

12. What could be done to mitigate this worst-case memory access latency?

A cache, indexed on virtual addresses, storing recently accessed data

Both cache and TLB could improve average memory access latency

A translation lookaside buffer (TLB) storing recently accessed page lookups

Neither cache nor TLB â€“ instead increase the depth of the page table to two levels

13. What is the purpose of the NX bit in the page table protection metadata?

States that this memory address is not accessible by the process

States that this memory address contains shared data that is not exclusive to this process

States that this memory address does not contain executable code for the process

States that this memory address can be read, but not written by the process

Look at the following memory configuration, based on an 8-bit byte-addressable virtual address space with single- level paging. In each 8-bit virtual address, the most significant byte refers to the page number and the least significant byte refers to the offset within the page.

Consider the following low-level C memory copy operations, involving the routine
void memcpy(void dst, void src, size_t n)
which copies n bytes from address src to address dst. With reference to the above page table, answer the questions below.

14. To which frame(s) does memcpy(0x10, 0x02, 4) write data?

frame 2 only

none â€“ an error is triggered

no frames

frames 2 and 3

15. To which frame(s) does memcpy(0x10, 0x00, 32) write data?

frames 2 and 3

none â€“ an error is triggered

frame 2 only

frame 7 only

16. To which frame(s) does memcpy(0x30, 0x00, 1) write data?

frame 2 only

an error is triggered

frames 2, 3 and 7

frame 3 only

17. To which frame(s) does memcpy(0x20, 0x30, 16) write data?

frame 2 only

an error is triggered

frame 3 only

frames 2 and 3

18. In practice, why is it unlikely that 8-bit memory would feature a virtual addressing scheme?

Such a small system is likely to be a single-process embedded system

There wouldnâ€™t be enough space for page table metadata

Mooreâ€™s law means 8-bit machines are no longer viable

Virtual memory causes unacceptable slowdown

19. Which of the following concepts is not used in the File Allocation Table (FAT) storage system?

Data Cluster

File Allocation Table

Root Directory

Logging History Database

20. Suppose a disk has X clusters. How many bits will be required for a cluster address in FAT directory entries?

2^X â€“ i.e. 2 raised to power X

ceiling (log_2 (X))

floor (log_2 (X))

21. Why are FAT filesystems restricted to fixed length filename extensions of three characters?

This is a convention imposed by DOS, rather than an actual physical restriction

FAT runs on 32 bit machines, so we cannot fit more than three characters into a machine word

There is a fixed number of bytes for the extension in FAT directory entry metadata

Developer convenience, since 26^3 is approximately equal to 2^16

22. What happens if a file is too large to fit into one cluster in a FAT file system?

The file must be placed on consecutive clusters on the disk, so it is contiguous

The FAT table has a pointer for each cluster to the address of the next cluster, so the file is laid out in a linked list format

The directory has an array of cluster addresses for each file

There is a maximum file size, so files are constrained to fit in a single cluster

23. On a FAT-formatted disk, suppose every file is N bytes in size, but the cluster size is M bytes, where M > N. What is the proportion of wasted disk space in the limit, due to internal fragmentation?

1 â€“ (N/M)

M/N

M - N

N/M

24. Which of the following is a genuine status for a FAT data cluster, indicated by the file allocation table?

Colossal

Corrupted

Consecutive

Compressed

25. Why is it efficient to locate a large file in contiguous disk clusters?

To reduce disk seek time

All of the above

To reduce internal fragmentation

To reduce size of file metadata in tables

26. What is the advantage of having a duplicate file allocation table for the file system?

Redundancy, in case of corruption in tables

Parallel access improves file access times

All of the above

It supports staged updates and undo operations for file allocation

27. Which of the following is an advantage of Unix inodes over the FAT structure?

Inodes enable more efficient small files than FAT

An inode's array of direct access blocks means we can seek past the first sector more quickly

Inodes enable much larger file sizes than FAT

Inodes support more efficient metadata access and modification

28. Why arenâ€™t read-only filesystems like data CDs formatted using FAT?

More space on disk means we can afford a larger directory structure

Data is statically laid out and never going to change on a read-only file system.

Because FAT is specific to a proprietary operating system

Seek times are lower for optical drives like CDs

Operating Systems (H) - 2017 Exam

Section A

1. How many characters 'C' are going to be printed?

2. How many characters 'D' are going to be printed?

3. Which is the minimum number of characters 'A' that will be printed?

4. Is the sequence of the characters: 'CABABDDCABCABD' possible to be printed (with exactly this order)?

Now consider the following cooperating processes A and B that share the (common) variable X:

The initial value of X is 5 (X=5;) before any process execution. The commands within each process are executed sequentially.

5. How many different values can the variable X get after the execution of both processes? Select one of the following:

Let us modify both programs by adding a semaphore S, which is initialized to 1 (S=1;), and X is initially 5 (X=5;). That is:

6. How many different values can the variable X get after the execution of both processes? Select one of the following:

Let us modify again the initial programs by adopting a semaphore T, which is initialized to 0 (T=0;), and X is initially 5 (X=5;). That is:

7. How many different values can the variable X get after the execution of both processes? Select one of the following:

Assume there are processes P and Q with the semaphores s and t, both initialized to 1 (s=1; t=1). Which of the following sequences of the commands P1, ..., P8, Q1, ..., Q5, leads those processes to a deadlock state, if any?

8. Select one of the following:

Section B

9. How many bytes of memory will there be?

10. For this system, consider using a virtual addressing scheme with single-level paging. If each page contains 16 bytes, how many pages will there be?

11. In the worst case, what happens to memory access latency in a virtual addressing environments with a single-level page table, with respect to physical addressing?

12. What could be done to mitigate this worst-case memory access latency?

13. What is the purpose of the NX bit in the page table protection metadata?

14. To which frame(s) does memcpy(0x10, 0x02, 4) write data?

15. To which frame(s) does memcpy(0x10, 0x00, 32) write data?

16. To which frame(s) does memcpy(0x30, 0x00, 1) write data?

17. To which frame(s) does memcpy(0x20, 0x30, 16) write data?

18. In practice, why is it unlikely that 8-bit memory would feature a virtual addressing scheme?

Section C

19. Which of the following concepts is not used in the File Allocation Table (FAT) storage system?

20. Suppose a disk has X clusters. How many bits will be required for a cluster address in FAT directory entries?

21. Why are FAT filesystems restricted to fixed length filename extensions of three characters?

22. What happens if a file is too large to fit into one cluster in a FAT file system?

23. On a FAT-formatted disk, suppose every file is N bytes in size, but the cluster size is M bytes, where M > N. What is the proportion of wasted disk space in the limit, due to internal fragmentation?

24. Which of the following is a genuine status for a FAT data cluster, indicated by the file allocation table?

25. Why is it efficient to locate a large file in contiguous disk clusters?

26. What is the advantage of having a duplicate file allocation table for the file system?

27. Which of the following is an advantage of Unix inodes over the FAT structure?

28. Why arenâ€™t read-only filesystems like data CDs formatted using FAT?

Operating Systems (H) - 2017 Exam

Section A

1. How many characters 'C' are going to be printed?

2. How many characters 'D' are going to be printed?

3. Which is the minimum number of characters 'A' that will be printed?

4. Is the sequence of the characters: 'CABABDDCABCABD' possible to be printed (with exactly this order)?

Now consider the following cooperating processes A and B that share the (common) variable X: The initial value of X is 5 (X=5;) before any process execution. The commands within each process are executed sequentially.

5. How many different values can the variable X get after the execution of both processes? Select one of the following:

Let us modify both programs by adding a semaphore S, which is initialized to 1 (S=1;), and X is initially 5 (X=5;). That is:

6. How many different values can the variable X get after the execution of both processes? Select one of the following:

Let us modify again the initial programs by adopting a semaphore T, which is initialized to 0 (T=0;), and X is initially 5 (X=5;). That is:

7. How many different values can the variable X get after the execution of both processes? Select one of the following:

Assume there are processes P and Q with the semaphores s and t, both initialized to 1 (s=1; t=1). Which of the following sequences of the commands P1, ..., P8, Q1, ..., Q5, leads those processes to a deadlock state, if any?

8. Select one of the following:

Section B

9. How many bytes of memory will there be?

10. For this system, consider using a virtual addressing scheme with single-level paging. If each page contains 16 bytes, how many pages will there be?

11. In the worst case, what happens to memory access latency in a virtual addressing environments with a single-level page table, with respect to physical addressing?

12. What could be done to mitigate this worst-case memory access latency?

13. What is the purpose of the NX bit in the page table protection metadata?

14. To which frame(s) does memcpy(0x10, 0x02, 4) write data?

15. To which frame(s) does memcpy(0x10, 0x00, 32) write data?

16. To which frame(s) does memcpy(0x30, 0x00, 1) write data?

17. To which frame(s) does memcpy(0x20, 0x30, 16) write data?

18. In practice, why is it unlikely that 8-bit memory would feature a virtual addressing scheme?

Section C

19. Which of the following concepts is not used in the File Allocation Table (FAT) storage system?

20. Suppose a disk has X clusters. How many bits will be required for a cluster address in FAT directory entries?

21. Why are FAT filesystems restricted to fixed length filename extensions of three characters?

22. What happens if a file is too large to fit into one cluster in a FAT file system?

23. On a FAT-formatted disk, suppose every file is N bytes in size, but the cluster size is M bytes, where M > N. What is the proportion of wasted disk space in the limit, due to internal fragmentation?

24. Which of the following is a genuine status for a FAT data cluster, indicated by the file allocation table?

25. Why is it efficient to locate a large file in contiguous disk clusters?

26. What is the advantage of having a duplicate file allocation table for the file system?

27. Which of the following is an advantage of Unix inodes over the FAT structure?

28. Why arenâ€™t read-only filesystems like data CDs formatted using FAT?

Now consider the following cooperating processes A and B that share the (common) variable X:

The initial value of X is 5 (X=5;) before any process execution. The commands within each process are executed sequentially.