Operating System Virtual Memory Virendra Singh Associate Professor

Operating System Virtual Memory Virendra Singh Associate Professor Computer Architecture and Dependable Systems Lab Department of Electrical Engineering Indian Institute of Technology Bombay http: //www. ee. iitb. ac. in/~viren/ E-mail: viren@ee. iitb. ac. in EE-717/453 Advanced Computing for Electrical Engineers Lecture 38

Benefits of Virtual Memory • Illusion of very large physical memory • Protection: by providing each process a separate virtual address space – Page table mechanism prevents one process from accessing the memory of another – Hardware is able to write-protect designated pages • Sharing – Common code (libraries, editors, etc. ) – For shared memory communication 07 Nov 2012 EE-717/EE-453@IITB 2

OS Responsibilities in VM Systems • Maintain/update page tables – Action initiated by addition of new process to memory or by a page fault • • Maintain list of free pages Execute page-replacement algorithms Write “dirty” pages back to disk Sometimes, update TLB (TLB may be either hardware or software managed) 07 Nov 2012 EE-717/EE-453@IITB 3

Page Replacement • Page replacement algorithms are needed when a page fault occurs and memory is full. • Use information collected by hardware to try to guess which pages are no longer in use • Most common approach: some variation of Least Recently Used (LRU), including various Clock – Keep pages that have been referenced recently on the assumption that they are in the current locality • Free frame pool is replenished periodically. 07 Nov 2012 EE-717/EE-453@IITB 4

Page Replacement - comments • Locality of reference is weaker today due mainly to OO programming – Lots of small functions & objects – Dynamic allocation/deallocation of objects on heap is more random than allocation on stack. • OS generally maintains a pool of free frames and uses them for both page replacement and for the file system cache. This also affects requirements for replacement. 07 Nov 2012 EE-717/EE-453@IITB 5

Page Replacement Algorithms • Page fault forces choice – which page must be removed – make room for incoming page • Modified page must first be saved – unmodified just overwritten • Better not to choose an often used page – will probably need to be brought back in soon 07 Nov 2012 EE-717/EE-453@IITB 6

Optimal Page Replacement Algorithm • Replace page needed at the farthest point in future – Optimal but unrealizable • Estimate by … – logging page use on previous runs of process – although this is impractical 07 Nov 2012 EE-717/EE-453@IITB 7

Not Recently Used Page Replacement Algorithm • Each page has Reference bit, Modified bit – bits are set when page is referenced, modified • Pages are classified 1. 2. 3. 4. not referenced, not modified not referenced, modified referenced, not modified referenced, modified • NRU removes page at random – from lowest numbered non empty class 07 Nov 2012 EE-717/EE-453@IITB 8

FIFO Page Replacement Algorithm • Maintain a linked list of all pages – in order they came into memory • Page at beginning of list replaced • Disadvantage – page in memory the longest may be often used 07 Nov 2012 EE-717/EE-453@IITB 9

Second Chance Page Replacement Algorithm • Operation of a second chance – pages sorted in FIFO order – Page list if fault occurs at time 20, A has R bit set (numbers above pages are loading times) 07 Nov 2012 EE-717/EE-453@IITB 10

Least Recently Used (LRU) • Assume pages used recently will used again soon – throw out page that has been unused for longest time • Must keep a linked list of pages – most recently used at front, least at rear – update this list every memory reference !! • Alternatively keep counter in each page table entry – choose page with lowest value counter – periodically zero the counter 07 Nov 2012 EE-717/EE-453@IITB 12

Simulating LRU in Software (1) LRU using a matrix – pages referenced in order 0, 1, 2, 3, 2, 1, 0, 3, 2, 3 07 Nov 2012 EE-717/EE-453@IITB 13

Simulating LRU in Software (2) • The aging algorithm simulates LRU in software • Note 6 pages for 5 clock ticks, (a) – (e) 07 Nov 2012 EE-717/EE-453@IITB 14

The Working Set Page Replacement Algorithm • The working set is the set of pages used by the k most recent memory references • w(k, t) is the size of the working set at time, t 07 Nov 2012 EE-717/EE-453@IITB 15

The Working Set Page Replacement Algorithm The working set algorithm 07 Nov 2012 EE-717/EE-453@IITB 16

The WSClock Page Replacement Algorithm 07 Nov 2012 EE-717/EE-453@IITB 17

Review of Page Replacement Algorithms 07 Nov 2012 EE-717/EE-453@IITB 18

Design Issues for Paging Systems Local versus Global Allocation Policies (1) • How memory should be allocated among the competing runnable processes ? • Original configuration • Local page replacement • Global page replacement 07 Nov 2012 EE-717/EE-453@IITB 23

Design Issues for Paging Systems Local versus Global Allocation Policies (2) • Global algorithm works better – Working set size vary over life time of process • Local algorithm – Size grows thrashing – Sizes shrinks memory wastage • Page allocation – Working set – Equal space to all processes – Process size - dynamic 07 Nov 2012 EE-717/EE-453@IITB 24

Local versus Global Allocation Policies (3) Page fault rate as a function of the number of page frames assigned 07 Nov 2012 EE-717/EE-453@IITB 25

Load Control • Despite good designs, system may still thrash • When PFF algorithm indicates – some processes need more memory – but no processes need less • Solution : Reduce number of processes competing for memory – swap one or more to disk, divide up pages they held – reconsider degree of multiprogramming 07 Nov 2012 EE-717/EE-453@IITB 26

Page Size (1) Small page size • Advantages – less internal fragmentation – better fit for various data structures, code sections – less unused program in memory • Disadvantages – programs need many pages, larger page tables 07 Nov 2012 EE-717/EE-453@IITB 27

Page Size (2) • Overhead due to page table and internal fragmentation page table space internal fragmentation • Where – s = average process size in bytes – p = page size in bytes – e = page entry 07 Nov 2012 EE-717/EE-453@IITB Optimized when 28

Separate Instruction and Data Spaces • One address space • Separate I and D spaces 07 Nov 2012 EE-717/EE-453@IITB 29

Shared Pages Two processes sharing same program sharing its page table 07 Nov 2012 EE-717/EE-453@IITB 30

Cleaning Policy • Need for a background process, paging daemon – periodically inspects state of memory • When too few frames are free – selects pages to evict using a replacement algorithm • It can use same circular list (clock) – as regular page replacement algorithmbut with diff ptr 07 Nov 2012 EE-717/EE-453@IITB 31