Scribd
Upload
92 views

Paging

The document discusses paging in operating systems. It explains that the OS manages physical memory addresses so that each process sees itself at the beginning of memory, simplifying memory management. When a process needs more memory than is available, the OS performs swapping by moving pages to disk to free up space for other processes. Common algorithms for determining which pages to swap out include first-in, first-out (FIFO) and least recently used (LRU). The document also briefly discusses interrupts and how they allow hardware devices like keyboards to notify the CPU of events without wastefully polling.

Uploaded by

vipul_sharma_50
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF, TXT or read online on Scribd
92 views

Paging

The document discusses paging in operating systems. It explains that the OS manages physical memory addresses so that each process sees itself at the beginning of memory, simplifying memory management. When a process needs more memory than is available, the OS performs swapping by moving pages to disk to free up space for other processes. Common algorithms for determining which pages to swap out include first-in, first-out (FIFO) and least recently used (LRU). The document also briefly discusses interrupts and how they allow hardware devices like keyboards to notify the CPU of events without wastefully polling.

Uploaded by

vipul_sharma_50
Copyright
© Attribution Non-Commercial (BY-NC)
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 7

Paging - operating systems resides at the beginning of memory - OS know exactly how much physical memory there is - OS can

address memory by its actual physical address (non-translated) - processes running under an OS cannot see physical memory addresses - address translated by the OS result: each process sees itself - at the beginning of memory - as the only process in the memory this greatly simplifies memory management from the programs view, can start addressing at #00000000 and OS takes care of translation - each process sees more memory that it actually has (because it thinks its the only one running) - swap provides extra space (more on this later) - capacity management is handled by the OS Examples (on Linux, diagrams by Prof. Mann) For simp licity, assume kernel = OS

No processes are loaded, but the translation already set up, user programs see memory as empty, except the part taken by the OS, which appears at the end of available memory.

One process loaded, sees itself at the beginning, but in actuality loaded right after the kernel.

Some more processes

No room for the Emacs process to load fully. It loads some, then swaps out Bash (which was accessed last) to disk, and loads the rest. Even though the memory space taken by the process is split, translation allows the process to think its still all together.

This allows each process to have its own memory space, thus uncontrolled behavior can be contained. Uncontrolled behavior by the operating system, on the other hand, makes for data loss and other mayhem, since it can access everything directly. Now the details: which page to swap out? We got several possibilities First In First Out (FIFO) - Store age of each page - the page that has been in memory longest is the one offloaded Good: - straightforward to implement - requiredslittle effort by the system to maintain age information (can just keep a counter, record and increment when page loaded) Bad: - say a page that has been loaded first is in heavy use - it will be offloaded anyways, and will have to be reloaded almost right away, so performance suffers Optimizing FIFO (Second Chance) - Every time a page is used, set a signal bit to 1 - When trying to swap out, if bit is 0, swap it out, if 1, set to 0 and try next one Good: - frequently used pages will have the bit set to 1, and hopefully will not get swapped out Bad: - if all pages are in heavy use, system will have to try every single one - some performance penalty to update the bit Least Recently Used (LRU) - record last time the page was used, i.e. update timestamp on each access Good: - in theory, more optimal than FIFO Bad: - often slows down the system so much, its not worth it (can increase memory access time by factor of 5) Real Life: UNIX pagedaemon uses a somewhat modified FIFO Second Chance.

Interrupts (hardware) - 15 (or so) signal lines used by hardware to tell CPU they need attention - Example: keyboard trivial implementation of keyboard: - ask every few ms if a key is pressed wasteful, can miss keystroked - this is known as polling, very wasteful CPU use with interrupts: - keyboard takes the keypress, and remembers it - sets the interrupt, so the CPU can get the data asap

References: http://www.wearcam.org/ece385/lecture7/pagedmemory.htm http://home.lanet.lv/~sd70058/aboutos/node99.html

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy