Demand Paging

Question

Can anyone pls explain what is the concept applied here

http://www.cs.jhu.edu/~yairamir/cs418/os6/sld010.htm

Comments

Hello Aashish.

I doubt the correctness of this answer ans explanation.

i can try to explain what i know.

EAT=(1-P_pf )*T_MAT + P_pf(P_{dirty bit} *T_{if replaced page is modified} +(1-P_{dirty bit} )T _{page fault time})

Now i will try to explain how to calculate EAT(Effective memory access time).

When there is no page fault then We directly access the main memory and find the data there(I am not considering TLB case hope you can take care of that like first check if data in TLB if yes then okay otherwise go to memory).

Now when there is page fault then there can we two cases

1) Page that is not there , we fetch that page from Hard disk but we have to make sure that there are free pages to place that page in main memory , if there are free frames available , its good otherwise we have to make free space explicitly by replacing some frame (Page replacement algorithms)

2) Now as we have to replace page then there are again two cases .let suppose you are replacing page x with page y (y is currently needed) then we have to remove x but before removing x make sure if it was updated or not and if updated then the updated copy saved on disk or not.

tell me if you have some doubt or i made some argument wrong in your consideration.

---

you can read this thread https://stackoverflow.com/questions/18550370/calculate-the-effective-access-time

---

what is meaning of page fault overhead.can anyone explain