Memory Management

It is the task carried out by the OS and hardware to accommodate multiple processes in main
memory. If only a few processes can be kept in main memory, then much of the time all
processes will be waiting for I/O and the CPU will be idle. Hence, memory needs to be allocated
efficiently in order to pack as many processes into memory as possible. In most schemes, the
kernel occupies some fixed portion of main memory and the rest is shared by multiple processes.

Memory Management Requirements

 Relocation
o Programmer cannot know where the program will be placed in memory when it is
executed.
o a process may be (often) relocated in main memory due to swapping
o swapping enables the OS to have a larger pool of ready-to-execute processes
o memory references in code (for both instructions and data) must be translated to
actual physical memory address
 Protection
o processes should not be able to reference memory locations in another process
without permission
o impossible to check addresses at compile time in programs since the program
could be relocated
o address references must be checked at run time by hardware
 Sharing
o must allow several processes to access a common portion of main memory without
compromising protection
 cooperating processes may need to share access to the same data structure
 better to allow each process to access the same copy of the program rather than
have their own separate copy
 Logical Organization
o users write programs in modules with different characteristics
 instruction modules are execute-only
 data modules are either read-only or read/write
 some modules are private others are public
o To effectively deal with user programs, the OS and hardware should support a basic form
of module to provide the required protection and sharing
 Physical Organization
o secondary memory is the long term store for programs and data while main memory
holds program and data currently in use
o moving information between these two levels of memory is a major concern of memory
management (OS)
 it is highly inefficient to leave this responsibility to the application programmer

Simple Memory Management

In this lecture, we study the simpler case where there is no virtual memory. An executing process
must be loaded entirely in main memory (if overlays are not used). Although the following
simple memory management techniques are not used in modern OS, they lay the ground for a
proper discussion of virtual memory:
o fixed partitioning
o dynamic partitioning
o simple paging
o simple segmentation

Fixed Partitioning
 Partition main memory into a set of non-overlapping regions called partitions
 Partitions can be of equal or unequal sizes
 any process whose size is less than or equal to a partition size can be loaded into the
partition
 if all partitions are occupied, the operating system can swap a process out of a partition
 a program may be too large to fit in a partition. The programmer must then design the
program with overlays: when the module needed is not present the user program must load
that module into the program’s partition, overlaying whatever program or data are there
 Main memory use is inefficient. Any program, no matter how small, occupies an entire
partition. This is called internal fragmentation.
 Unequal-size partitions lessens these problems but they still remain...
 Equal-size partitions was used in early IBM’s OS/MFT (Multiprogramming with a Fixed
number of Tasks)

Placement Algorithm with Partitions

 Equal-size partitions
o If there is an available partition, a process can be loaded into that partition
 because all partitions are of equal size, it does not matter which partition is
used
o If all partitions are occupied by blocked processes, choose one process to swap
out to make room for the new process
 Unequal-size partitions: use of multiple queues
o assign each process to the smallest partition within which it will fit
o A queue for each partition size
o tries to minimize internal fragmentation
o Problem: some queues will be empty if no processes within a size range is present
 Unequal-size partitions: use of a single queue
o When its time to load a process into main memory the smallest available partition
that will hold the process is selected
o increases the level of multiprogramming at the expense of internal fragmentation

Dynamic Partitioning
 Partitions are of variable length and number
 Each process is allocated exactly as much memory as it requires
 Eventually holes are formed in main memory. This is called external fragmentation
 Must use compaction to shift processes so they are contiguous and all free memory is in
one block
 Used in IBM’s OS/MVT (Multiprogramming with a Variable number of Tasks)

Dynamic Partitioning: an example

 A hole of 64K is left after loading 3 processes: not enough room for another process
 Eventually each process is blocked. The OS swaps out process 2 to bring in process 4
 another hole of 96K is created
 Eventually each process is blocked. The OS swaps out process 1 to bring in again process
2 and another hole of 96K is created...
 Compaction would produce a single hole of 256K

Placement Algorithm
 Used to decide which free block to allocate to a process
 Goal: to reduce usage of compaction (time consuming)
 Possible algorithms:
o Best-fit: choose smallest hole
o First-fit: choose first hole from beginning
o Next-fit: choose first hole from last placement

Placement Algorithm: comments

 Next-fit often leads to allocation of the largest block at the end of memory
 First-fit favors allocation near the beginning: tends to create less fragmentation then
Next-fit
 Best-fit searches for smallest block: the fragment left behind is small as possible
o main memory quickly forms holes too small to hold any process: compaction
generally needs to be done more often

Replacement Algorithm
 When all processes in main memory are blocked, the OS must choose which process to
replace
o A process must be swapped out (to a Blocked-Suspend state) and be replaced by a
new process or a process from the Ready-Suspend queue
o We will discuss later such algorithms for memory management schemes using
virtual memory

Buddy System
 A reasonable compromise to overcome disadvantages of both fixed and variable
partitioning schemes
 A modified form is used in Unix SVR4 for kernel memory allocation
 Memory blocks are available in size of 2K where L <= K <= U and where
o 2L = smallest size of block allocatable
o 2U = largest size of block allocatable (generally, the entire memory available)
 We start with the entire block of size 2U
 When a request of size S is made:
o If 2U-1 < S <= 2U then allocate the entire block of size 2U
o Else, split this block into two buddies, each of size 2U-1
 o If 2U-2 < S <= 2U-1 then allocate one of the 2 buddies
o Otherwise one of the 2 buddies is split again
 This process is repeated until the smallest block greater or equal to S is generated
 Two buddies are coalesced whenever both of them become unallocated
 The OS maintains several lists of holes
o the i-list is the list of holes of size 2i
o whenever a pair of buddies in the i-list occur, they are removed from that list and
coalesced into a single hole in the (i+1)-list
 Presented with a request for an allocation of size k such that 2i-1 < k <= 2i:
o the i-list is first examined
o if the i-list is empty, the (i+1)-list is then examined...

Example of Buddy System

 
Buddy Systems: remarks
 On average, internal fragmentation is 25%
o each memory block is at least 50% occupied
 Programs are not moved in memory
o simplifies memory management
 Mostly efficient when the size M of memory used by the Buddy System is a power of 2
o M = 2U "bytes" where U is an integer
o then the size of each block is a power of 2
o the smallest block is of size 1
o Ex: if M = 10, then the smallest block would be of size 5

Relocation
 Because of swapping and compaction, a process may occupy different main memory
locations during its lifetime
 Hence physical memory references by a process cannot be fixed
 This problem is solved by distinguishing between logical address and physical address

Address Types
 A physical address (absolute address) is a physical location in main memory
 A logical address is a reference to a memory location independent of the physical
structure/organization of memory
 Compilers produce code in which all memory references are logical addresses
 A relative address is an example of logical address in which the address is expressed as a
location relative to some known point in the program (ex: the beginning)

Address Translation
 Relative address is the most frequent type of logical address used in pgm modules (ie:
executable files)
 Such modules are loaded in main memory with all memory references in relative form
 Physical addresses are calculated "on the fly" as the instructions are executed
  For adequate performance, the translation from relative to physical address must by done
by hardware

Simple example of hardware translation of addresses

 When a process is assigned to the running state, a base register (in CPU) gets loaded with
the starting physical address of the process
 A bound register gets loaded with the process’s ending physical address
 When a relative addresses is encountered, it is added with the content of the base register
to obtain the physical address which is compared with the content of the bound register
 This provides hardware protection: each process can only access memory within its
process image

Simple Paging
 Main memory is partition into equal fixed-sized chunks (of relatively small size)
 Trick: each process is also divided into chunks of the same size called pages
 The process pages can thus be assigned to the available chunks in main memory called
frames (or page frames)
 Consequence: a process does not need to occupy a contiguous portion of memory

Example of process loading

 Now suppose that process B is swapped out
 When process A and C are blocked, the pager loads a new process D consisting of 5
pages
 Process D does not occupied a contiguous portion of memory
 There is no external fragmentation
 Internal fragmentation consist only of the last page of each process

Page Tables
 The OS now needs to maintain (in main memory) a page table for each process
 Each entry of a page table consist of the frame number where the corresponding page is
physically located
 The page table is indexed by the page number to obtain the frame number
 A free frame list, available for pages, is maintained

Logical address used in paging

 Within each program, each logical address must consist of a page number and an offset
within the page
 A CPU register always holds the starting physical address of the page table of the
currently running process
 Presented with the logical address (page number, offset) the processor accesses the page
table to obtain the physical address (frame number, offset)

Logical address in paging

 The logical address becomes a relative address when the page size is a power of 2
  Ex: if 16 bits addresses are used and page size = 1K, we need 10 bits for offset and have
6 bits available for page number
 Then the 16 bit address obtained with the 10 least significant bit as offset and 6 most
significant bit as page number is a location relative to the beginning of the process
 By using a page size of a power of 2, the pages are invisible to the programmer,
compiler/assembler, and the linker
 Address translation at run-time is then easy to implement in hardware
o logical address (n,m) gets translated to physical address (k,m) by indexing the
page table and appending the same offset m to the frame number k
Simple Segmentation
 Each program is subdivided into blocks of non-equal size called segments
 When a process gets loaded into main memory, its different segments can be located
anywhere
 Each segment is fully packed with instructs/data: no internal fragmentation
 There is external fragmentation; it is reduced when using small segments
 In contrast with paging, segmentation is visible to the programmer
o provided as a convenience to organize logically programs (ex: data in one
segment, code in another segment)
o must be aware of segment size limit
 The OS maintains a segment table for each process. Each entry contains:
o the starting physical addresses of that segment.
o the length of that segment (for protection)

Logical address used in segmentation

 When a process enters the Running state, a CPU register gets loaded with the starting
address of the process’s segment table.
 Presented with a logical address (segment number, offset) = (n,m), the CPU indexes (with
n) the segment table to obtain the starting physical address k and the length l of that
segment
 The physical address is obtained by adding m to k (in contrast with paging)
o the hardware also compares the offset m with the length l of that segment to
determine if the address is valid

Simple segmentation and paging comparison

 Segmentation requires more complicated hardware for address translation
 Segmentation suffers from external fragmentation
 Paging only yield a small internal fragmentation
 Segmentation is visible to the programmer whereas paging is transparent
 Segmentation can be viewed as commodity offered to the programmer to organize
logically a program into segments and using different kinds of protection (ex: execute-
only for code but read-write for data)
o for this we need to use protection bits in segment table entries