Competing For Memory

1
Competing For Memory

• Vivek Pai / Kai Li
• Princeton University

2
Mechanics

• Feedback optionally anonymous
• No real retribution anyway
• Do it to make me happy
• Quiz 1 Question 2 answer(s)
• regs ! bits
• Registers at top of memory hierarchy
• Lots of acceptable answers
• Last Quiz, Feedback still being digested

3
The Big Picture

• Weve talked about single evictions
• Most computers are multiprogrammed
• Single eviction decision still needed
• New concern allocating resources
• How to be fair enough and achieve good overall
  throughput
• This is a competitive world local and global
  resource allocation decisions

4
Lessons From Enhanced FIFO

• Observations
• its easier to evict a clean page than a dirty
  page
• sometimes the disk and CPU are idle
• Optimization when systems free, write dirty
  pages back to disk, but dont evict
• Called flushing often falls to pager daemon

5
x86 Page Table Entry
Page frame number
D
L
Gl
Cw
P
U
A
Cd
Wt
O
W
V
12
31

• Valid
• Writable
• Owner (user/kernel)
• Write-through
• Cache disabled
• Accessed (referenced)
• Dirty
• PDE maps 4MB
• Global

Reserved
6
Program Behaviors

• 80/20 rule
• gt 80 memory references are made by lt 20 of code
• Locality
• Spatial and temporal
• Working set
• Keep a set of pages in memory would avoid a lot
  of page faults

Working set
page faults
pages in memory
7
Observations re Working Set

• Working set isnt static
• There often isnt a single working set
• Multiple plateaus in previous curve
• Program coding style affects working set
• Working set is hard to gauge
• Whats the working set of an interactive program?

8
Working Set

• Main idea
• Keep the working set in memory
• An algorithm
• On a page fault, scan through all pages of the
  process
• If the reference bit is 1, record the current
  time for the page
• If the reference bit is 0, check the last use
  time
• If the page has not been used within d, replace
  the page
• Otherwise, go to the next
• Add the faulting page to the working set

9
WSClock Paging Algorithm

• Follow the clock hand
• If the reference bit is 1, set reference bit to
  0, set the current time for the page and go to
  the next
• If the reference bit is 0, check last use time
• If page has been used within d, go to the next
• If page hasnt been used within d and modify bit
  is 1
• Schedule the page for page out and go to the next
• If page hasnt been used within d and modified
  bit is 0
• Replace this page

10
Simulating Modify Bit with Access Bits

• Set pages read-only if they are read-write
• Use a reserved bit to remember if the page is
  really read-only
• On a read fault
• If it is not really read-only, then record a
  modify in the data structure and change it to
  read-write
• Restart the instruction

11
Implementing LRU without Reference Bit

• Some machines have no reference bit
• VAX, for example
• Use the valid bit or access bit to simulate
• Invalidate all valid bits (even they are valid)
• Use a reserved bit to remember if a page is
  really valid
• On a page fault
• If it is a valid reference, set the valid bit and
  place the page in the LRU list
• If it is a invalid reference, do the page
  replacement
• Restart the faulting instruction

12
Demand Paging

• Pure demand paging relies only on faults to bring
  in pages
• Problems?
• Possibly lots of faults at startup
• Ignores spatial locality
• Remedies
• Loading groups of pages per fault
• Prefetching/preloading
• So why use it?

13
Speed and Sluggishness

• Slow is gt .1 seconds (100 ms)
• Speedy is ltlt .1 seconds
• Monitors tend to be 60 Hz
• lt16.7ms between screen paints
• Disks have seek rotational delay
• Seek is somewhere between 7-16 ms
• At 7200rpm, one rotation 1/120 sec 8ms.
  Half-rotation is 4ms
• Conclusion? One disk access OK, six are bad

14
Memory Pressure

• Swap space
• Region of disk used to hold overflow
• Contains only data pages (stack/heap/globals).
  Why?
• Swap may exist as regular file, but dedicated
  region of disk more common

15
Disk Address

• Use physical memory as a cache for disk
• Where to find a page on a page fault?
• PPage field is a disk address
• Observation OS knows that pages are real but not
  in memory

Virtual address space
Physical memory
invalid
16
Imagine a Global LRU

• Global across all processes
• Idea when a page is needed, pick the oldest
  page in the system
• Problems? Process mixes?
• Interactive processes
• Active large-memory sweep processes
• Mitigating damage?

17
Source of Disk Access

• VM System
• Main memory caches - full image on disk
• Filesystem
• Even here, caching very useful
• New competitive pressure/decisions
• How do we allocate memory to these two?
• How do we know were right?

18
Partitioning Memory

• Originally, specified by administrator
• 20 used as filesystem cache by default
• On fileservers, admin would set to 80
• Each subsystem owned pages, replaced them
• Observation theyre all basically pages
• Why not let them compete?
• Result unified memory systems file/VM

19
File Access Efficiency

• read(fd, buf, size)
• Buffer in processs memory
• Data exists in two places filesystem cache
  processs memory
• Known as double buffering
• Various scenarios
• Many processes read same file
• Process wants only parts of a file, but doesnt
  know which parts in advance

20
Result Memory-Mapped Files
Process A
Process B
Process C
Process A
Process B
Process C
File
Map
File
Map
File
Map
File
21
Lazy Versus Eager

• Eager do things right away
• read(fd, buf, size) returns bytes read
• Bytes must be read before read completes
• What happens if size is big?
• Lazy do them as theyre needed
• mmap() returns pointer to mapping
• Mapping must exist before mmap completes
• When/how are bytes read?
• What happens if size is big?

22
Semantics How Things Behave

• What happens when
• Two process obtain data (read or mmap)
• One process modifies data
• Two processes obtain data (read or mmap)
• A third process modifies data
• The two processes access the data

23
Being Too Smart

• Assume a unified VM/File scheme
• Youve implemented perfect Global LRU
• What happens on a filesystem dump?

24
Amdahls Law

• Gene Amdahl (IBM, then Amdahl)
• Noticed the bottlenecks to speedup
• Assume speedup affects one component
• New time
• (1-not affected) affected/speedup
• In other words, diminishing returns

25
NT x86 Virtual Address Space Layouts
00000000
Application code Globals Per-thread stacks DLL
code
3-GB user space
7FFFFFFF 80000000
Kernel exec HAL Boot drivers
C0000000 C0800000
Process page tables Hyperspace
BFFFFFFF C0000000
System cache Paged pool Nonpaged pool
1-GB system space
FFFFFFFF
FFFFFFFF
26
Virtual Address Space in Win95 and Win98
00000000
User accessible
Unique per process (per application), user mode
7FFFFFFF 80000000
Shared, process-writable (DLLs, shared
memory, Win16 applications)
Systemwide user mode
C0000000
Win95 and Win98
Systemwide kernel mode
Operating system (Ring 0 components)
FFFFFFFF
27
Details with VM Management

• Create a processs virtual address space
• Allocate page table entries (reserve in NT)
• Allocate backing store space (commit in NT)
• Put related info into PCB
• Destroy a virtual address space
• Deallocate all disk pages (decommit in NT)
• Deallocate all page table entries (release in NT)
• Deallocate all page frames

28
Page States (NT)

• Active Part of a working set and a PTE points to
  it
• Transition I/O in progress (not in any working
  sets)
• Standby Was in a working set, but removed. A
  PTE points to it, not modified and invalid.
• Modified Was in a working set, but removed. A
  PTE points to it, modified and invalid.
• Modified no write Same as modified but no write
  back
• Free Free with non-zero content
• Zeroed Free with zero content
• Bad hardware errors

29
Dynamics in NT VM
Demand zero fault
Page in or allocation
Standby list
Free list
Zero list
Bad list
Process working set
Modified writer
Zero thread
Soft faults
Modified list
Working set replacement
30
Shared Memory

• How to destroy a virtual address space?
• Link all PTEs
• Reference count
• How to swap out/in?
• Link all PTEs
• Operation on all entries
• How to pin/unpin?
• Link all PTEs
• Reference count

w
. . .
. . .
Page table
. . .
Process 1
w
Physical pages
. . .
. . .
Page table
Process 2
31
Copy-On-Write

• Childs virtual address space uses the same page
  mapping as parents
• Make all pages read-only
• Make child process ready
• On a read, nothing happens
• On a write, generates an access fault
• map to a new page frame
• copy the page over
• restart the instruction

r
r
. . .
. . .
Page table
. . .
Parent process
r
r
Physical pages
. . .
. . .
Page table
Child process
32
Issues of Copy-On-Write

• How to destroy an address space
• Same as shared memory case?
• How to swap in/out?
• Same as shared memory
• How to pin/unpin
• Same as shared memory