Chapter Four

Device Management

12/31/2023 School of Computing, DDUIoT 1

 Contents
o overview
o Principles of I/O hardware
o I/O devices
o Devices controller
o Memory-mapped I/O
o Direct memory access(DMA)
o Interrupt revisited
o Principle of I/O software
o Goals of I/O and issues of software
o I/O operation
o I/O software layers
o Disks

12/31/2023 School of Computing, DDUIoT 2

 Overview
o The control of device connected to computer is major
concern of an operating system.
o Operating system issue commands to the devices, catch
interrupts, and handle errors.
o It should also provide an interface between the devices
and the rest of the system that is simple and easy to
use.
o Because I/O devices varied so widely in their function
and speed, varied method is needed to control them.
o These method form the I/O subsystem of the kernel.

12/31/2023 School of Computing, DDUIoT 3

 I/O devices
o OS concerns with the way data are handled by the I/O
device.
o There are two types of I/O devices: blocked and character.
 Block devices
– Stores information in Fixed-size blocks, each one with
its own address.
– Common block size range from 512bytes to 32,768bytes.
– All transfers are in units of one or more entire
(consecutive) blocks.
– The essential property of a block device is that it is
possible to read or write each block independently of
all the other ones.
• Example: Hard disks, CD-ROMs, and USB sticks.
12/31/2023 School of Computing, DDUIoT 4
 I/O devices(con't…)
 Character devices
– Delivers or accepts stream of character, without
regard to any block structure.
– It is not addressable and does not have any seek
operation.
o Example: Printers, network interfaces, mice (for
pointing)
o Most other devices that are not disk-like can be seen
as character devices.
o Some devices just do not fit into both above
classification.
o Clocks, for example, are not block addressable, Nor
do they generate or accept character streams.
12/31/2023 School of Computing, DDUIoT 5
 Device controller
 Is collection of electronics that can operate a port, bus
or devices
 For example disk controller’s job is to convert the serial
bit stream into a block of bytes and perform any error
correction necessary.
 Controller has a few registers that are used for
communicating with the CPU.
 By writing into these registers, the operating system can
command the device to deliver data, accept data, switch
itself on or off, or otherwise perform some action.
 By reading from these registers, the operating system
can learn what the device's state is, whether it is
prepared to accept a new command, and so on.

12/31/2023 School of Computing, DDUIoT 6

 Memory-mapped I/O
 How the CPU communicates with controllers to accomplish I/O transfer?
Two alternatives exists.
1. Use of special I/O instruction (IN and OUT):
 Each register is assigned an I/O port number, an 8- or 16-bit integer.
 The set of all the I/O ports form the I/O port space
 Protected ordinary user programs cannot access it (only the operating
system can).
Example:
IN REG, PORT: the CPU can read in control register PORT and store the
result in CPU register REG.
OUT PORT, REG: the CPU can write the contents of REG to a control
register.
2. Mapping controller register to memory space (Memory- Mapped I/O):
 Each register is assigned a unique memory address to which no
memory is assigned.
 CPU executes I/O requests using the standard data transferring
instruction to read and write controller registers.
12/31/2023 School of Computing, DDUIoT 7
 Direct memory access(DMA)
o Requesting data from an I/O controller one byte at a
time wastes the CPU's time.
o So a different scheme, called DMA (Direct Memory
Access) chip that can control the flow of bits between
memory and some controller without constant CPU
intervention.
o The CPU sets up the DMA chip, telling it how many
bytes to transfer, the device and memory addresses
involved, and the direction, and lets it go.
o When the DMA chip is done, it causes an interrupt.

12/31/2023 School of Computing, DDUIoT 8

 Interrupt revisited
o When an i/o device has finished the work given to it, it
causes an interrupt
• It does this by asserting a signal on a bus line that it
has been assigned.
• This signal is detected by the interrupt controller chip
on the parent board, which then decides what to do.
• If no other interrupts are pending, the interrupt
controller processes the interrupt immediately.
• If another one is in progress, or another device has
made a simultaneous request on a higher-priority
interrupt request line on the bus, the device is just
ignored for the moment.
• In this case it continues to assert an interrupt signal on
the bus until it is serviced by the CPU.
12/31/2023 School of Computing, DDUIoT 9
 Interrupt revisited (cont….)
o To handle the interrupt:
– The controller puts a number on the address lines specifying
which device wants attention and asserts a signal to interrupt
the CPU.
– New program counter is fetched from table called
interrupt vector using the number on the address
lines is used as an index into this table.
– Interrupt service procedure is start running and
finally acknowledges the interrupt by writing a
certain value to one of the interrupt controller's I/0
ports.

12/31/2023 School of Computing, DDUIoT 10

 Goals and issues of i/o software
o Device Independence:
– It should be possible to write programs that can access any i/o
device without having to specify the device in advance.
– It is up to the operating system to take care of the problems
caused by the fact that these devices are really different.
o Uniform Naming:
₋ The name of a file or a device should simply be a string or an
integer and not dependent on the device in any way.
o Error handling:
- In general, errors should be handled as close to the hardware as
possible (at the device controller level).
- Many errors are transient, such as read errors caused by specks
of dust on the read head, and will go away if the operations are
repeated.

12/31/2023 School of Computing, DDUIoT 11

 Goals and issues of i/o software
o TransferThere are two types of transfer modes :
Synchronous (Blocking) and Asynchronous (interrupt –driven).
– Synchronous transfer: the program requesting I/O transfer will
be suspended until the transfer is completed.
– Asynchronous transfer: the CPU starts the transfer and goes off
to do something until the interrupt that shows the completion
of the transfer arrives.

Device types There are two device types
Shareable (such as disk)
 Used by many users at the same time.
 No problems are caused by multiple users having open files on the
same disk at the same time.
Dedicated (tape drives)
 Have to be dedicated to a single user until that user is finished.
 Having two or more users writing blocks intermixed at random to the
same tape will definitely not work.
12/31/2023 School of Computing, DDUIoT 12
 I/O Operation |Programmed I/O
 There are three fundamentally different ways that I/O
operation can be performed.
 Programmed I/O
 Interrupt-driven I/O
 I/O using DMA
 Programmed I/O
 is a method in which the main CPU inputs or outputs
each byte or word and sits in a tight loop waiting until it
can get or send the next one.
 The CPU continuously polls the device to see if it is ready
or not. This behavior is often called polling or busy
waiting.
 Programmed I/O is simple but has the disadvantage of
12/31/2023 tying up the CPU School
fulloftime DDUIoT all the I/0 is done .
until
Computing, 13
 I/O Operation |Interrupt-driven I/O
 Interrupt-driven I/O
 is method in which the CPU starts an I/O transfer
for a character or word and goes off to do
something else until an interrupt arrives signaling
completion of the I/O.
 The advantage of interrupt-driven I/O is allow the
CPU to do something else while waiting for the
device to become ready for the next work.
 The disadvantage of interrupt-driven I/0 is that an
interrupt occurs on every character or word which
Interrupts take time, that wastes a certain amount
of CPU time.

12/31/2023 School of Computing, DDUIoT 14

I/O Operation |Direct memory access(DMA)
 I/O using DMA
 the way in which a separate chip manages the complete
transfer of a block of data, given an interrupt only when the
entire block has been transferred.
 DMA is programmed I/O, only with the DMA controller doing
all the work, instead of the main CPU.
 This strategy requires special hardware (the DMA controller)
but frees up the CPU during the I/O to do other work.
 The big win with DMA is reducing the number of interrupts
from one per character to one per buffer.
 The DMA controller is usually much slower than the main
CPU. If the DMA controller is not capable of driving the
device at full speed, or the CPU usually has nothing to do any
way while waiting for the DMA interrupt, then interrupt-
driven I/0 or even programmed I/0 may be better.
12/31/2023 School of Computing, DDUIoT 15
 Direct Memory Access(DMA)

Operation of DMA tranfer

12/31/2023 School of Computing, DDUIoT 16

 I/0 software layers
o I/0 software is typically organized in four layers.
o Each layer has a well-defined function to perform and
a well-defined interface to the adjacent layers.
o The functionality and interfaces differ from system to
system.

12/31/2023 School of Computing, DDUIoT 17

 Interrupt handlers
o Interrupts are unpleasant facts of program or process
and cannot be avoided.
o They should be hidden away deep in the bowels of the
operating system.
o The best way to hide them is to have driver starting an
I/O operation block until the I/O has completed and the
interrupt occurs.
o When the interrupt happens, the interrupt procedure
does whatever it has to do in order to handle the
interrupt. Then unblock the driver that started it.

12/31/2023 School of Computing, DDUIoT 18

 Device driver
o All devices – Dependent code goes in the device driver.
o Each device driver handles one device type, or at most,
one class of closely related devices.
o The job of a device driver is to accept requests from
the device independent software above it and sees to it
that the request is executed.
o Steps in carrying out I/O requests
– Translate it from abstract to concrete terms.
– Write into the controller’s device registers.
– The device driver blocks itself until interrupt comes.
– The blocked driver will be awakened by the interrupt.
– Device driver checks for errors.
– If everything is all right, the driver may have data to pass to
the device independent software.
–
12/31/2023If nothing is queued, the
School of driver
Computing, DDUIoTblocks waiting for the next
19
request
 Device independent I/O software
o It is large fraction of I/O software.
o Perform the I/0 functions that are common to all
devices and to provide a uniform interface to the user-
level software.
o Functions of the device-independent I/O software:
 Uniform interfacing for device drivers – Perform I/O
function common to all drives.
 Device Naming – Responsible for mapping symbolic
devices names onto the proper driver.
 Device protection – Prevent users from accessing
devices that they are not entitled to access.
 Providing device-independent block size – Provide
uniform block size to higher layers hiding differences in
block sizes.
12/31/2023 School of Computing, DDUIoT 20
 Device independent I/O software
 Buffering: If a user process write half a block, the OS
will normally keep the data in buffer until the rest of
the data are written. Keyboard inputs that arrive before
it is needed also require buffering.
 Allocating and releasing dedicated devices: It is up to
the OS to examine requests for devices usage and
accept or reject them.
 Error reporting: Errors handling, by and large, is done
by drivers. Most errors are device dependent. After the
driver tries to read the block a certain number of
times, it gives up and informs the device-independent
software. It then reports to the caller.

12/31/2023 School of Computing, DDUIoT 21

 User level I/O software
 Although most of the i/o software is within the
operating system, a small portion of it consists of
libraries linked together with user programs.
 System calls, including the I/O system calls, are
normally made by library procedures.
 The collection of all these library procedures is clearly
part of the I/0 system.
 Not all user-level I/0 software consists of library
procedures.
 Another important category is the spooling system
which is a way of dealing with dedicated I/O devices in
a multi-programming system.

12/31/2023 School of Computing, DDUIoT 22

 Summary of I/O software layers

12/31/2023 School of Computing, DDUIoT 23

 Disk
o All real disks are organized into cylinders/platter, each
platter is arranged like a record, each one containing
many tracks.
o Each of the tracks then will be divided into sectors (equal
number of sectors or different number of sectors).
o In the case of equal number of sectors
– The data density as closer to the center (hub) is high.
– The speed increases as the read/write moves to the outer tracks.
o Modern large hard drives have more sectors per track on
outer tracks e.g. IDE drives.
o Many controllers can read or write on one drive while
seeking on one or more other drives, but floppy disk
controller cannot do that.

12/31/2023 School of Computing, DDUIoT 24

 Disk (con’t…)

The ugly guts of a hard disk.

12/31/2023 School of Computing, DDUIoT 25
 Disk access time
o The time required to read or write a disk block is
determined by three factors
o The seek time: measures the delay for the disk head to
reach the right track.
o The rotational delay: accounts for the time to get to the
right sector.
o Transfer time: is how long the actual data read or write
takes.
o There may be additional over head for the operating
system or the controller hardware on the hard disk drive.
o Rotational speed: measured in revolutions per minute or
RPM, partially determines the rotational delay and transfer
time.
o Manufacturers often report average seek times: These times
average the time to seek from any track to any other
12/31/2023 School of Computing, DDUIoT 26
track.
 Disk access time(con’t…)
o Rotational delay depends partly on how fast the disk
platters spin.
o Average rotational delay = 0.5 x rotations x rotational speed
o For example, a 5400 RPM disk has an average rotational delay of:
0.5 rotations /(5400 rotations/minute) = 5.55ms.
o The overall response time is the sum of the seek time,
rotational delay, transfer time, and overhead.
 Example: Assume a disk has the following
specifications.
 An average seek time of 9ms
 A 5400 RPM rotational speed
 A 10MB/s average transfer rate
 2ms of overheads

12/31/2023 School of Computing, DDUIoT 27

 Disk access time(con’t…)
How long does it take to read a random 1,024 byte
sector?
 The average rotational delay is 5.55ms.
 The transfer time will be about (1024 bytes / 10
MB/s) = 0.1ms.
 So the response time is then 9ms + 5.55ms + 0.1ms +
2ms = 16.7ms.
 That’s 16,700,000 cycles for a 1GHz processor!

12/31/2023 School of Computing, DDUIoT 28

 Disk scheduling algorism
o the seek time dominates the other two times, so
reducing the mean seek time can improve system
performance substantially.
o There is different disk scheduling algorithm is used to
reduce mean seek time.
 First Come First Served (FCFS)
 Shortest Seek Time First (SSTF)
 SCAN (Elevator) Algorithm
 C-SCAN (Modified Elevator) Algorithm (Circular-
SCAN)

12/31/2023 School of Computing, DDUIoT 29

 First Come First Served (FCFS)
o Accept a request one at a time and carries them out
in that order.
o Service requests in the order they are received.
o The simplest and the fairest of all, but it doesn’t
improve performance
E.g: Track initial position: 11
 Track request: 1,36,16,34,9,12
 Service order: 1,36,16,34,9,12

12/31/2023 School of Computing, DDUIoT 30

 Shortest Seek Time First (SSTF)
o It handles the closest (the least disk arm movement)
request next, to minimize seek time.
o Performance (efficiency): provides better performance.
o Possibility of starvation: it lacks fairness
Example: Track initial position: 11
 Track request: 1,36,16,34,9,12
 Service order: 12,9,16, 1, 34, 36

12/31/2023 School of Computing, DDUIoT 31

 SCAN (Elevator) Algorithm
o The disk arm keeps moving in the same direction until
there are no more outstanding requests in that
direction, and then it switches direction.
o It has the head start at track 0 and move towards the
highest numbered track, servicing all requests for a
track as it passes the track.
o SCAN algorithm is guaranteed to service every request
in one complete pass through the disk.
o It provides better service distribution.
Example. Track initial position: 11
 Track request: 1,36,16,34,9,12
 Direction bit: 1
 Service order: 12, 16, 34, 36, 9, 1
12/31/2023 School of Computing, DDUIoT 32
 C-SCAN (Modified Elevator) Algorithm (Circular-SCAN)
o A slight modification of elevator algorithm that has
smaller variance in response times is always scans in the
same direction .
o When the highest numbered cylinder with a pending
request has been serviced, the arm goes to the lowest-
numbered cylinder with a pending request and then
continues moving in an upward direction.
o In effect, the lowest-numbered cylinder is thought of as
being just above the highest-numbered cylinder.
o It reduces the maximum delay experienced by new
request.
Example: Track initial position: 11
 Track request: 1,36,16,34,9,12
 Direction bit: 1
 Service order: 12, 16, 34, 36, 1, 9
12/31/2023 School of Computing, DDUIoT 33
 Redundant array independent disk(RAID)
o Redundant array of independent disks may be used to
increase disk reliability.
o In a RAID system, a single large file is stored in several
separate disk units by breaking the file up into a number
of smaller pieces and storing these pieces on different
disks.
o When a file is accessed for a read, all disks deliver their
data in parallel.
o RAID may be implemented in hardware or in the
operating system.

12/31/2023 School of Computing, DDUIoT 34

12/31/2023 School of Computing, DDUIoT 35