Sie sind auf Seite 1von 2

HW 7

1.

(5 points) The total size is equal to the number of entries times the size of each entry.
Each page is 16KB, and thus, 14 bits of the virtual and physical address will be used as
a page offset. The remaining 40 14 = 26 bits of the virtual address constitute the
virtual page number, and there are thus 226 entries in the page table, one for each virtual
page number. Each entry requires 36 14 = 22 bits to store the physical page number
and an additional 4 bits for the valid, protection, dirty, and use bits. We round the 26
bits up to a full word per entry, so this gives us a total size of 226 32 bits or 256MB.

2.

(30 points)
a) Virtual address 64 bits, physical memory 16 GB
Page size 4 KB or 12 bits, page table entry 8 bytes or 3 bits
#PTE = 64 12 = 52 bits or 252 entries
PT physical memory = 252 23 bytes = 255 bytes
b) Virtual address 64 bits, physical memory 16 GB
Page size 4 KB or 12 bits, page table entry 8 bytes or 3 bits
#PTE = 64 12 = 52 bits or 252 entries
4 KB page/8 byte PTE = 29 pages indexed per page
Hence with 252 PTEs will need 6-level page table setup.
Each address translation will require at least 6 physical memory accesses.

c) Virtual address 64 bits, physical memory 16 GB


page size 4 KB or 12 bits, page table entry 8 bytes or 3 bits
#PTE = 64 12 = 52 bits or 252 entries
Since there are only 16GB physical memory, only 2(3412) PTEs are really needed to
store the page table.
Common-case: no hash conflict, so one memory reference per address translation;
Worst case: almost 2(3412) memory references are needed if hash table degrade into a
link list.
d) TLB initialization, or process context switch.
e) TLB miss. When most missed TLB entry is cached in processor caches.
f) Write protection exception.
3.

(15 points)
a) System A requires 20TB (10 TB for mirroring), System B requires 12.5TB (additional
2.5TB for parity). There System A requires 7.5TB more storage than System B.

b) System A worst case will take 2 x 30ms, as data write and mirror write is performed
one after the after. System B worst case takes 4 x 30ms (Read old data, read old parity,
write new data, write new parity).
c) For System B, 1 disk can fail before data loss occurs. Best case for System A 4 disks
can fail before data loss occurs, worst case for system A 1 disk can fail before data
loss.
4.

Each transaction needs 5 x 10,000 = 50,000 instructions


CPU is maxed out @ 500,000,000/50,000 = 10,000 transactions per second
A I/O limit = 1,500/5 = 300 transactions per second
B I/O limit = 1,000/5 = 200 transactions per second
Therefore the I/Os are the bottlenecks limiting speed, i.e. they are the maximum
sustainable speed.

Das könnte Ihnen auch gefallen