619 views
1 votes
1 votes

1 Answer

1 votes
1 votes
@Gabbar. Consider this one. I resolve the faults at that level itself as Arjun does did for one of the questions in GO17.

That means page fault occurs at M.M level, resolve it there itself. If page fault  occurs at instruction level, resolve it at instruction level.

Average M.M access time

= page hit ratio * ( M.M acccess time ) +

   pag fault ratio * ( M.M access time to discover page fault + page fault service time)

= 0.8 * (300 ns) + 0.2 * (300 + 500)

= 400 ns

Now, average instruction execution time

= CPU time + page access time

= CPU time + TLB hit ratio * (TLB access time + 1 M.M access)

  + TLB miss ratio * ( TLB access to discover TLB doesnt have the entry + 3 M.M accesses + load the entry in TLB)

= 200 + 0.8 *( 1 M.M access) + 0.2 * (3 M.M accesses)........................TLB access time negligible

= 200 + 0.8(400) + 0.2*(3 * 400)

= 760 ns

Related questions

0 votes
0 votes
0 answers
2
Senthilkar asked Aug 11, 2018
172 views
Consider the system with four type Of resource. Which of the following statstateme is true
0 votes
0 votes
0 answers
3
3 votes
3 votes
1 answer
4
papesh asked Nov 13, 2016
306 views
i know there is deadlock, what about the starvation??