More on Address Translation CS170 Fall 2015. T. Yang Based on - - PowerPoint PPT Presentation

more on address translation
SMART_READER_LITE
LIVE PREVIEW

More on Address Translation CS170 Fall 2015. T. Yang Based on - - PowerPoint PPT Presentation

Misc Exercise 2 updated with Part III. Due on next Tuesday 12:30pm. Project 2 (Suggestion) Write a small test for each call. Start from file system calls with simple Linux translation. Or start from Exec() Midterm (May


slide-1
SLIDE 1

Misc

  • Exercise 2 updated with Part III.
  • Due on next Tuesday 12:30pm.
  • Project 2 (Suggestion)
  • Write a small test for each call.
  • Start from file system calls with simple

Linux translation.

  • Or start from Exec()
  • Midterm (May 7)
  • Close book. Bring 3 pages of double-

sided notes.

  • Next Monday: Project 2 or review?
slide-2
SLIDE 2

More on Address Translation

CS170 Fall 2015. T. Yang Based on Slides from John Kubiatowicz http://cs162.eecs.Berkeley.edu

slide-3
SLIDE 3

Implementation Options for Page Table

  • Page sharing among process
  • What can page table entries be

utilized?

  • Page table implementation
  • One-level page table
  • Muti-level paging
  • Inverted page tables
slide-4
SLIDE 4

Shared Pages through Paging

  • Shared code
  • One copy of read-only code shared among

processes (i.e., text editors, compilers, window systems).

  • Shared code must appear in same location in the

logical address space of all processes

  • Private code and data
  • Each process keeps a separate copy of the code

and data

  • The pages for the private code and data can

appear anywhere in the logical address space

slide-5
SLIDE 5

Shared Pages Example

slide-6
SLIDE 6

PageTablePtrB

page #0 page #1 page #2 page #3 page #5

V,R N V,R,W N

page #4

V,R V,R,W

page #4

V,R

Example

Offset Virtual Page #

Virtual Address (Process A): PageTablePtrA

page #0 page #1 page #3 page #4 page #5

V,R V,R

page #2

V,R,W V,R,W N V,R,W

Offset Virtual Page #

Virtual Address (Process B):

Shared Page

This physical page appears in address space of both processes

page #2

V,R,W

slide-7
SLIDE 7

Optimization of Unix System Call Fork()

  • A child process copies address space of parent.
  • Most of time it is wasted as the child performs exec().
  • Can we avoid doing copying on a fork()?

PageTablePtrB

page #0 page #1 page #2 page #3 page #5

V,R N V,R,W N

page #4

V,R V,R,W

page #4

V,R

Offset Virtual Page #

Virtual Address (Process A): PageTablePtrA

page #0 page #1 page #3 page #4 page #5

V,R V,R

page #2

V,R,W V,R,W N V,R,W

Offset Virtual Page #

Virtual Address (Child proc B):

Parent address space

page #2

V,R,W

Child address space

slide-8
SLIDE 8

Unix fork() optimization

PageTablePtr

page #0 page #1 page #2 page #3 page #5

V,R N V,R,W N

page #4

V,R V,R,W

page #4

V,R

Offset Virtual Page #

Virtual Address (Process A): PageTablePtrA

page #0 page #1 page #3 page #4 page #5

V,R V,R

page #2

V,R,W V,R,W N V,R,W

Offset Virtual Page #

Virtual Address (Child proc):

Parent address space

page #2

V,R,W

Child address space

slide-9
SLIDE 9

Unix fork() optimization

PageTablePtr

page #0 page #1 page #2 page #3 page #5

V,R N V,R,W N

page #4

V,R V,R,W

page #4

V,R

Offset Virtual Page #

Virtual Address (Process A): PageTablePtrA

page #0 page #1 page #3 page #4 page #5

V,R V,R

page #2

V,R,W V,R,W N V,R,W

Offset Virtual Page #

Virtual Address (Child proc):

Parent address space

page #2

V,R,W

Child address space

Child

slide-10
SLIDE 10

Copy-on-Write: Lazy copy during process creation

  • COW allows both parent and child processes to

initially share the same pages in memory.

  • A shared page is duplicated only when modified
  • COW allows more efficient process creation as only

modified pages are copied

slide-11
SLIDE 11

Copy on Write: After Process 1 Modifies Page C

How to memorize a page is shared? When to detect the need for duplication? Need a page table entry bit Physical page number Page table entry

slide-12
SLIDE 12

More examples of utilizing page table entries

  • How do we use the PTE?
  • Invalid PTE can imply different things:

– Region of address space is actually invalid or – Page/directory is just somewhere else than memory

  • Validity checked first

– OS can use other bits for location info

  • Usage Example: Copy on Write
  • Indicate a page is shared with a parent
  • Usage Example: Demand Paging
  • Keep only active pages in memory
  • Place others on disk and mark their PTEs invalid

Physical page number Page table entry

slide-13
SLIDE 13

Example: Intel x86 architecture PTE

  • Address format (10, 10, 12-bit offset)
  • Intermediate page tables called “Directories”

P: Present (same as “valid” bit in other architectures) W: Writeable U: User accessible PWT: Page write transparent: external cache write-through PCD: Page cache disabled (page cannot be cached) A: Accessed: page has been accessed recently D: Dirty (PTE only): page has been modified recently L: L=14MB page (directory only). Bottom 22 bits of virtual address serve as offset

Page Frame Number (Physical Page Number) Free (OS) 0 L D A PCD

PWT

U W P 1 2 3 4 5 6 7 8 11-9 31-12

slide-14
SLIDE 14

More examples of utilizing page table entries

  • Usage Example: Zero Fill On Demand
  • Security and performance advantages

– New pages carry no information

  • Give new pages to a process initially with PTEs marked as

invalid. – During access time, page fault  physical frames are allocated and filled with zeros

  • Often, OS creates zeroed pages in background
  • Can a process modify its own translation tables?
  • NO!
  • If it could, could get access to all of physical memory
  • Has to be restricted
slide-15
SLIDE 15

Implementation Options for Page Table

  • Page sharing among process
  • What can page table entries be

utilized?

  • Page table implementation
  • One-level page table
  • Muti-level paging
  • Inverted page tables
slide-16
SLIDE 16

One-Level Page Table

  • What is the maximum size of

logical space?

Mapping of pages Constraint: Each page table needs to fit into a physical memory page! Why? A page table needs consecutive space. Memory allocated to a process is a sparse set of nonconsecutive pages Maximum size = # entry * page size

slide-17
SLIDE 17

One-level page table cannot handle large space

  • Example:
  • 32 -bit address space with 4KB per page.
  • Page table would contain 232/ 212= 1 million

entries. – 4 bytes per entry

  • Need a 4MB page table with contiguous space.

– Is there 4MB contiguous space for each process?

  • Maximum size with 4KB per page

#entry= 4KB/4B = 1K. Maximum logical space=1K*4KB= 4MB.

slide-18
SLIDE 18

One-level Page Table: Advantage/Disadvantage

  • Pros
  • Simple memory allocation
  • Easy to Share
  • Con: What if address space is sparse?
  • E.g. on UNIX, code starts at 0, stack starts at

(231-1).

  • Cannot handle a large virtual address space
  • Con: What if table really big?
  • Not all pages used all the time  would be nice

to have working set of page table in memory

  • How about combining paging and

segmentation?

  • Segments with pages inside them?
  • Need some sort of multi-level translation
slide-19
SLIDE 19

Two-Level Page-Table Scheme

Level 1 Level 2

slide-20
SLIDE 20

Address-Translation Scheme

Level 1 Level 2 p1 is an index into the outer level-1 page table, p2 is the displacement within the page of the level-2 inner page table Level-2 page table gives the final physical page ID

slide-21
SLIDE 21

Physical Address:

Offset Physical Page #

4KB

Flow of the two-level page table

10 bits 10 bits 12 bits Virtual Address:

Offset Virtual P2 index Virtual P1 index 4 bytes

PageTablePtr

  • Tree of Page Tables
  • Tables fixed size (1024 entries)
  • On context-switch: save single

PageTablePtr register

  • Valid bits on Page Table Entries
  • Don’t need every 2nd-level table
  • Even when exist, 2nd-level tables

can reside on disk if not in use

4 bytes

slide-22
SLIDE 22

stack

Summary: Two-Level Paging

1111 1111

stack heap code data

Virtual memory view 0000 0000 0100 0000 1000 0000 1100 0000 page1 # offset Physical memory view

data code heap stack

0000 0000 0001 0000 0101 000 0111 000 1110 0000 page2 #

111 110 null 101 null 100 011 null 010 001 null 000 11 11101 10 11100 01 10111 00 10110 11 01101 10 01100 01 01011 00 01010 11 00101 10 00100 01 00011 00 00010 11 null 10 10000 01 01111 00 01110

Page Tables (level 2) Page Table (level 1) 1111 0000

slide-23
SLIDE 23

stack

Summary: Two-Level Paging

stack heap code data

Virtual memory view 1001 0000 (0x90) Physical memory view

data code heap stack

0000 0000 0001 0000 1000 0000 (0x80) 1110 0000

111 110 null 101 null 100 011 null 010 001 null 000 11 11101 10 11100 01 10111 00 10110 11 01101 10 01100 01 01011 00 01010 11 00101 10 00100 01 00011 00 00010 11 null 10 10000 01 01111 00 01110

Page Tables (level 2) Page Table (level 1)

slide-24
SLIDE 24

Constraint of paging

  • Bits of d = log (page size)
  • Bits of p1 >= log (# entries in level-1 table)
  • Bits of p2 >= log (# entries in level-2 table)
  • Physical page number is limited by entry size
  • f level-2 table.
  • logical space size = # entry in level-1 table *

# entry in level 2 table * page size

Level 1 Level 2 Physical page

slide-25
SLIDE 25

Analysis of a Two-Level Paging Example

  • A logical address (on 32-bit machine with

4K page size) is divided into:

  • a page number consisting of 20 bits
  • a page offset consisting of 12 bits
  • Each entry uses 4 bytes
  • How to build a two-level paging scheme?
  • How many entries can a single-page table

hold?

  • What are p1, p2?

page number page offset pi p2 d ? ? 12

slide-26
SLIDE 26

Analysis of a Two-Level Paging Example

  • A 1-page table with 4KB contains 1K

entries and each uses 4B.

  • 1K entries require 10 bits for P1 and P2
  • ffset
  • The page number is further divided into:
  • a 10-bit level-1 index
  • a 10-bit level-2 index

page number page offset pi p2 d 10 10 12 What if we use 2 bytes for each table entry?

  • Increased logical space size?
  • Increased physical space size?
slide-27
SLIDE 27

Example of maximum logical space size

  • Maximum logical space size

# entry in level-1 page table * # entry in level-2 page table * page size = 1K * 1K * 4KB = 232 bytes = 4GB

page number page offset pi p2 d 10 10 12 1K 1K 1K 1K 4K 4K

slide-28
SLIDE 28

An example of three-level paging in a 64-bit address space What is the maximum logical and physical space size?

slide-29
SLIDE 29

Three-level Paging in Linux

Address is divided into 4 parts

slide-30
SLIDE 30

Hashed Page Tables

  • Common in address spaces > 32 bits
  • Size of page table grows proportionally as large as

amount of virtual memory allocated to processes

  • Use hash table to limit the cost of search
  • to one — or at most a few — page-table entries
  • One hash table per process
  • This page table contains a chain of elements

hashing to the same location

  • Use this hash table to find the physical page of

each logical page

  • If a match is found, the corresponding physical

frame is extracted

slide-31
SLIDE 31

Hashed Page Table

slide-32
SLIDE 32

Inverted Page Table

  • One hash table for all processes
  • One entry for each real page of memory
  • Entry consists of the virtual address of

the page stored in that real memory location, with information about the process that owns that page

  • Decreases memory needed to store each

page table, but increases time needed to search the table when a page reference

  • ccurs
slide-33
SLIDE 33

Inverted Page Table Architecture

slide-34
SLIDE 34

Address Translation Comparison

Advantages Disadvantages Segmentation Fast context switching: Segment mapping maintained by CPU External fragmentation Paging (single-level page) No external fragmentation, fast easy allocation Large table size ~ virtual memory Internal fragmentation Paged segmentation Table size ~ # of pages in virtual memory, fast easy allocation Multiple memory references per page access Two-level pages Inverted Table Table size ~ # of pages in physical memory Hash function more complex

slide-35
SLIDE 35

Summary

  • Page Tables
  • Memory divided into fixed-sized chunks of memory
  • Virtual page number from virtual address mapped

through page table to physical page number

  • Offset of virtual address same as physical address
  • Large page tables can be placed into virtual memory
  • Usage of page table entries
  • Page sharing.
  • Copy on write
  • Pages on demand
  • Zero fill on demand
  • Multi-Level Tables
  • Virtual address mapped to series of tables
  • Permit sparse population of address space
  • Inverted page table
  • Size of page table related to physical memory size