10.4. Protection and Management in Page Tables

Now page as I told that if for a corresponding virtual corresponding virtual page number the physical page is not there is there is not a proper translation of the virtual page to a physical page, I have a page fault. What does that mean? I have loaded a virtual address, for that I have a virtual page number, the translation told me that corresponding to that virtual page number this virtual page does not currently reside in physical memory.

A page fault occurs when a program tries to access a page of memory that is not currently loaded in physical memory. This means that there is no valid translation from the virtual page number to a physical page. When this happens, the operating system needs to fetch the required page from secondary storage, such as a hard drive, to load it into a physical memory frame.

Imagine trying to read a book from a library, but the specific book you want is currently checked out. You'd need to wait for it to become available or go ask for it to be brought back from storage. This is similar to how page faults work in a computer system.

Page faults can be the page fault penalty for virtual memories is very high. Why because the access times on the secondary storage is very high. Now to access a, whereas let us say for accessing the main memory I will only take around say 50 to 70 nanoseconds for accessing the accessing the secondary storage I may take millions of nanoseconds.

The penalty for a page fault is significant due to the time it takes to access secondary storage. Accessing data from main memory takes only about 50 to 70 nanoseconds, while accessing data from secondary storage (like a hard disk) can take millions of nanoseconds. This large difference in access times results in high latency for programs that experience page faults.

Think of it like retrieving a file from a filing cabinet versus trying to retrieve it from a remote warehouse. The filing cabinet is quick to access, but if you have to go to the warehouse, it can take a long time, interrupting your work.

the page size that we decide like we have to decide what should be the size of a block of cache. Similarly we have to decide what should be the size of a page. Now page sizes should be large enough to amortize the high cost of accessing the secondary storage.

Choosing the right page size is crucial for optimizing memory access and reducing page faults. Pages need to be large enough to minimize the number of times the system must access slower secondary storage. Larger pages help in bringing more data into physical memory at once, which increases the chances of accessing required data without generating additional page faults.

It's like packing a suitcase for a trip. If you bring a larger suitcase, you can fit more items, reducing the number of trips you have to make back and forth to your home. Similarly, larger pages mean fewer trips to the 'warehouse' for data.

Typically today page sizes are of the order of 4 KB’s to 16 KB’s. The trend newer trends for desktops and servers are that it is going to be still higher to even say 32 KB’s or 64 KB’s.

Current operating systems typically use page sizes ranging from 4 KB to 16 KB, with newer systems moving towards even larger sizes like 32 KB or 64 KB. The reason for increasing page sizes is to further reduce the frequency of accessing secondary storage and minimize page faults in modern applications that require high performance.

If you think about how you store items, larger boxes can be more efficient. If you're packing for a picnic and have bigger containers, you can fit more food, which means fewer trips to the car to grab more items.

However, for embedded systems page sizes are typically lower of the order of 1 KB.

Embedded systems often use smaller page sizes, typically around 1 KB, because these systems are resource-constrained. Smaller pages help to reduce internal fragmentation, where unused space within the last page may lead to inefficient memory usage.

Think of a small-scale cafeteria where space is limited. If you use giant trays, you may waste a lot of space because not all the trays will be full. Using smaller trays helps to minimize waste and fits better in a compact environment.

Virtual memories are typically use fully associative placement of pages in main memory.

In virtual memory management, using a fully associative mapping allows any virtual page to be placed in any physical frame in memory. This flexibility reduces the chances of page faults, as the virtual memory management system can optimize placement based on current needs.

Consider how a movie theater seats its patrons. If they can seat anyone anywhere rather than locking them into specific seats (like reserved seats), they can optimize seating arrangements based on current attendance, which allows for better use of space.

However, because page faults are very expensive compared to that this the compared to that handling page faults in software is much lower in cost.

The operating system handles page faults through software mechanisms, which may be less efficient than hardware solutions but are more feasible given the high costs associated with page faults. Smart algorithms can be implemented in software to help reduce the frequency of page faults.

Imagine a restaurant kitchen. If chefs can anticipate a high demand for a certain dish, they can prepare more ingredients in advance to avoid delays. Similarly, smart algorithms help manage data loading effectively to minimize delays caused by page faults.

Suppose when I write on to a physical memory I don’t write to the to the corresponding location in virtual memory or the secondary storage. I use a write-back scheme because if I go on each time I write into physical memory if I have to write into secondary storage it will be hugely costly as we understand.

In a write-back memory system, updates to physical memory are not immediately mirrored in secondary storage. Instead, modifications are retained until the page needs to be replaced, at which point any 'dirty' pages—those that have been modified—are written back to secondary storage. This approach reduces the number of times data needs to be written to slower storage, improving performance.

Think of it like updating your contact information in a phonebook. You might make a lot of changes in your personal phone before you go back and update the actual phonebook at your home. This way, you avoid unnecessary trips back and forth, just like delaying updates to secondary storage.

Page table stores placement information it has an array of page table in entries indexed by virtual page number.

Each process has a page table that maps virtual pages to physical pages in memory. The entries in this table contain vital information like physical addresses and status indicators (e.g., presence in memory, dirty bits) that assist the operating system in managing memory access efficiently.

Imagine a library catalog that tracks where each book is located. If the catalog is organized and up-to-date, finding a book is fast and easy. Similarly, a well-organized page table allows the operating system to quickly translate virtual addresses to physical addresses.

--