Page 412 - FUNDAMENTALS OF COMPUTER

NPP

412 Fundamentals of Computers NPP

(iii) Interrupt Controlled Data Transfer (iii) B§Q´>ßQ> H§$Q´>moëS> S>mQ>m Q´>m§g\$a
The input-output device is started by the BZnwQ>-AmCQ>nwQ> {S>dmBg H$mo CPU Ûmam àma§^ {H$¶m
CPU and the CPU goes on executing its main OmVm h¡ VWm CPU BgHo$ ‘w»¶ àmoJ«m‘ H$mo E³Or³¶yQ> H$aZo
program. Whenever IO device is ready it gives Ho$ {bE OmVm h¡ O~ IO {S>dmBg V¡¶ma H$mo OmVr h¡ Vmo
a signal to the CPU over its interrupt line. When-
ever CPU is interrupted it completes the cur- ¶h BgH$s BÝQ>aßQ> bmBZ na CPU Ho$ {bE EH$ {g½Zb
rent instruction. Now the CPU starts execut- XoVr h¡ O~ H$^r CPU H$mo BÝQ>aßQ> {H$¶m OmVm h¡ Vmo ¶h
ing other program written for IO data transfer. dV©‘mZ BÝñQ´>³eZ H$mo nyam H$aVm h¡& A~ CPU IO S>mQ>m
This program is called Interrupt Service Rou- Q´>mÝg’$a Ho$ {bE {bIo J¶o Aݶ àmoJ«m‘ H$mo E³Or³¶yQ>
tine (ISR). Before branching to ISR the address H$aZm àma§^ H$aVm h¡& ¶h àmoJ«m‘ BÝQ>aßQ> g{d©g ê$Q>rZ
of next instruction is automatically push onto (ISR) H$hbmVm h¡& ISR ‘| ~«mqMJ Ho$ nhbo AJbr
the stack. Now data transfer takes place with BÝñQ´>³eZ H$m ES´>og ñdV… ñQ>oQ> na nyam hmo OmVm h¡& A~
the help of instructions written in ISR. After ISR ‘| {bIo J¶o BÝñQ´>³eÝg H$s ghm¶Vm go S>mQ>m Q´>m§g’$a
executing the ISR, the program control will hmoVm h¡& ISR H$mo E³Or³¶yQ> H$aZo Ho$ níMmV² àmoJ«m‘
automatically return to the main program and H$ÊQ´>mob ñdV… ‘w»¶ àmoJ«m‘ ‘| dmng Am OmVm h¡ VWm CPU
CPU will start executing the main program. ‘w»¶ àmoJ«m‘ H$mo E³Or³¶yQ> H$aoJm& Bg àH$ma, ¶h {d{Y
Thus this method gives efficient utilisation of CPU H$s Hw$eb Cn¶mo{JVm XoVr h¡& CPU H$m g‘¶ I/O
CPU. The time of CPU is not wasted in checking H$s ñQ>oQ>g H$s Om§M H$aZo ‘| Zï> Zht hmoVm h¡&
status of IO.

2. DMA Data Transfer 2. DMA S>mQ>m Q´>m§g\$a
The group of wires carrying information BÝ\$m°_}eZ bo OmZo dmbo dm`g© H$m g_yh {gñQ>_ ~g
is called System Bus. Normally, CPU is the bus H$hbmVm h¡Ÿ& gm_mÝ`V`m ~g na Omo ^r S>mQ>m Q´>m§g\$a
master. That means when data is to be trans- hmoVm h¡, dh CPU Ho$ {Z`§ÌU _| hmoVm h¡ Š`m|{H$ CPU
ferred between memory and the IO devices, it
must be under control of the CPU. But in some hr ~g H$m ñdm_r hmoVm h¡Ÿ& bo{H$Z `{X h_| A{YH$
data transfer the speed is reduced because of OmZH$mar Q´>m§g\$a H$aZm h¡ Am¡a H$moB© JUZm Zht H$aZm
the involvement of the CPU. Therefore a data h¡, Vmo EH$ Eogr VH$ZrH$ H$s Amdí`H$Vm hmoVr h¡ {Og_|
transfer technique is used in which IO devices CPU ~rM _| Z AmE Am¡a grYo-grYo _o_moar VWm BZnwQ>-
directly access the memory and CPU is by- AmCQ>nwQ> `wpŠV`m| _| S>mQ>m Q´>m§g\$a hmo& Eogr hr {d{Y H$mo
passed. This type of data transfer is called Di-
rect Memory Access (DMA) Data transfer. The DMA (S>m`aoŠQ> _o_moar EŠgog) S>mQ>m Q´>m§g\$a H$hVo
main advantage of DMA data transfer is the h¢Ÿ& Bg {d{Y _| CPU go {ZdoXZ H$aHo$ ~g H$m {Z`§ÌU
greater speed. It is useful when a bulk of data bo {b`m OmVm h¡Ÿ& CPU ~g H$mo CÀM à{VamoY H$s
is transferred and no computation is needed. AdñWm _| S>mb XoVm h¡ VWm grYo-grYo BZnwQ>-AmCQ>nwQ>
For example, data transfer between RAM and `wpŠV`m| d _o_moar Ho$ _Ü` S>mQ>m H$m Q´>m§g\$a hmoVm h¡ &
Floppy disk, Hard disk or high speed printer.
When you have prepared a program. You want BgH$m _w»` \$m`Xm `h h¡ {H$ `h EH$ A{YH$ J{V go
to save this program into hard disk. The pro- H$m`© H$aZo dmbm S>mQ>m Q´>m§g\$a hmoVm h¡Ÿ& O¡go, hmS>©
gram must be directly saved into Hard disk {S>ñH$, âbm°nr {S>ñH$ d ao_ Ho$ _Ü` S>mQ>m Q´>m§g\$a `{X