Operating Systems 2015F: Assignment 5: Difference between revisions
No edit summary |
No edit summary |
||
Line 11: | Line 11: | ||
#* a TCP/IP stack | #* a TCP/IP stack | ||
#* All of the above | #* All of the above | ||
# An OS that is based upon a microkernel | # An OS that is based upon a microkernel architecture will typically: | ||
architecture will typically: | |||
#* have most device drivers and filesystem code run within separate userspace processes | #* have most device drivers and filesystem code run within separate userspace processes | ||
#* have most OS code run in CPU supervisor mode | #* have most OS code run in CPU supervisor mode | ||
#* support loadable kernel modules | #* support loadable kernel modules | ||
#* All of the above | #* All of the above | ||
# A typical executing application is best | # A typical executing application is best described as a: | ||
described as a: | |||
#* kernel | #* kernel | ||
#* process | #* process |
Revision as of 05:11, 13 November 2015
This assignment is not yet finalized
- Within a UNIX filesystem, one inode can be referred to by many different pathnames (filenames). What is another term for these names?
- hard links
- symbolic links
- URLs
- all of the above
- A modern monolithic operating system kernel typically implements
- scheduling \& address space management
- some device drivers
- a TCP/IP stack
- All of the above
- An OS that is based upon a microkernel architecture will typically:
- have most device drivers and filesystem code run within separate userspace processes
- have most OS code run in CPU supervisor mode
- support loadable kernel modules
- All of the above
- A typical executing application is best described as a:
- kernel
- process
- thread
- system call
- When a running program X requests data from a file F whose contents are on disk, the OS will:
- save X's current state
- schedule a disk request for F's data blocks
- load the state of another ready-to-run program Y (which may be X)
- All of the above
- When a regular process is running on a single-CPU system, all of the following are true except:
- The kernel is running.
- Disks can be writing data to memory.
- The CPU will generate an exception (software interrupt) if certain addresses are accessed.
- The CPU is in user mode.
- Modern virtual memory systems:
- Logically divide RAM into fixed-sized chunks
- Store portions of processes on disk
- Allocate a logical address space to each process that is distinct from the physical memory map
- All of the above
- The ``dirty bit in a page table entry
indicates that:
- A security violation has occurred
- The corresponding page has been accessed.
- The corresponding page has been modified.
- The corresponding page is shared between multiple processes.
- An inode contains all of the following except:
- user ID
- filename
- time of last data modification
- link count
- When checking the consistency of a filesystem (e.g. after a loss of power), all of the following must be checked except:
- inode reference counts
- pointers to data blocks (in inodes or indirect blocks)
- contents of data blocks of regular files
- block free list
# If a system automatically boots into a defective kernel (e.g., one that crashes on startup), you can recover the system by:
- Select a different kernel in the bootloader
- Perform a soft reboot of the system
- Perform a hard reboot of the system
- Press Alt-SysRq-R
- Virtual memory is prohibitively expensive on a system that lacks a TLB because
- Every memory access then requires a system call.
- The page table has to be checked on every memory access
- The process table has to be updated on every tick interval.
- None of the above