Demand Paging

Demand Paging

• In demand paging, a page is loaded into primary memory only when the process references it. It is the simplest fetch policy implemented in virtual memory. Fig. 4.10.1 shows concept of demand paging.

• Demand paging guarantees that the system brings into primary memory only those pages that processes actually need. It involves interaction between hardware and software component of the virtual memory.

• Paging with swapping method will become as demand paging concept. Process's logical address space is stored on the secondary storage device.

• Demand paging requires that the page map table for each process keep track of each page as it is loaded or removed from primary memory..

• When user want to execute a process, user swap a process into a memory. Lazy swapper is used for swapping the process into memory. Lazy swapper swaps the pages whenever needed.

Working of demand paging

1. Program attempts to find a page.

2. If the page is located in the primary memory, then the program executes.

3. If the page cannot be found in the primary memory, then the page fault occurs.

4. Memory reference is checked for required page to determine it is a valid reference to a location on a secondary storage, if it is then the page will be required page. If not, the process is terminated.

5. Schedule disk operation to read the required page into primary memory.

6. Restart the instruction that was interrupted by the OS trap.

• Demand paging takes support from hardware to distinguish the pages. The valid-invalid bit method is used. This method gives the idea about page location. Page may be in the primary memory or in the secondary storage.

• If valid bit is set, then required page is stored in the primary memory. When an invalid bit is set, then page is on the secondary storage disk.

• Valid-invalid bit field is attached with page table. Fig. 4.10.2 shows page table entry.

• The valid-invalid bit field contain a boolean value to indicate whether the page exists in primary memory.

0 = Not in memory

1 = In memory

• Initially all valid-invalid entries are set to 0. This information is used for replacement algorithm. Page fault leads to the page replacement operation to load the required pages in primary memory.

A failure to find a page in memory is often called a page fault.

Demand paging is most efficient when users are aware of the page size used by operating system.

Steps in Handling a Page Fault

Fig. 4.10.3 shows page fault handling steps in demand paging.

When page fault occurs, the system performs following steps:

1. Check the process control block table for reference which bit is set i.e. valid bit or invalid bit for memory access.

2. If invalid bit is set, user will terminate the process. If valid bit is set, then the page is loaded into memory.

3. Check for free frame for new page loading.

4. Start reading secondary storage device for desired page for allocated frame.

5. Required page is loaded into the memory after completion of read operation. PCB table is updated.

6. System restart the instruction that was interrupted by the error. Page is loaded in memory so process can use it.

Demand Paging Performance

• Demand paging is a solution to inefficient memory utilization. Performance of the computer system is affected by demand paging. Effective access time for a demand paged memory is calculated as follows:

• Since demand paging like caching, can compute average access time. Memory access time (m) for most computers now ranges from 10 to 200 nanoseconds. If there is no page fault, then

Effective access time = Memory access time

• If there is page fault, then

Effective access time = Hit rate x Hit time + Miss rate x Miss time

= Hit time + Miss rate x Miss penalty

Effective access time = Memory access time * (1-p) +p* page fault service time

where p is the probability of a page fault

• Effective access time is directly proportional to page fault rate.

• Suppose memory access time = 100 ns, Page fault service time = 25 ms then Effective access time = Memory access time * (1 - p) + p * page fault service time = 100 (1 - p) + p * 25,000,000

= 100+ 24,999,900 p

Advantages of Demand Paging

1. Large virtual memory.

2. More efficient use of memory.

3. Unconstrained multiprogramming. There is no limit on degree of multiprogramming.

Disadvantages of Demand Paging

1. Number of tables and amount of processor over head for handling page interrupts are greater than in the case of the simple paged management techniques.

2. Due to the lack of an explicit constraints on a jobs address space size.

Comparison of Demand Paging with Segmentation