ImportSettings:

BaseSettings:BrownstoneDefault

HighestAnswerLetter:D

MultipleKeywordsinSameParagraph:No

Chapter:Chapter5

MultipleChoice

1A.xWhichofthefollowingistrueofcooperativescheduling?

zl

BvItrequiresatimer.

)

zAprocesskeepstheCPUuntilitreleasestheCPUeitherbyterminatingorbyswitchingto

由waitingstate.

ci)xe

zItincursacostassociatedwithaccesstoshareddata.

DIX

/Aprocessswitchesfromtherunningstatetothereadystatewhenaninterruptoccurs.

An

s:B

Feedback:5.1.3

Difficulty:Medium

z

A\isthenumberofprocessesthatarecompletedpertimeunit.

/!

B\CPUutilization

z)

c\Responsetime

zl

D\Turnaroundtime

7?Throughput

Ans:D

Feedback:5.2

Difficulty:Medium

3.schedulingisapproximatedbypredictingthenextCPUburstwithanexponential

averageofthemeasuredlengthsofpreviousCPUbursts.

A)Multilevelqueue

B)RR

C)FCFS

D)SJF

Ans:D

Feedback:5.3.2

Difficulty:Medium

4A.Theschedulingalgorithmisdesignedespeciallyfortime-sharingsystems.

\

J

ZSJF

BX

I

c/FCFS

x

)

zRR

D\

1

/Multilevelqueue

Ans:C

Feedback:5.3.4

Difficulty:Medium

5.Whichofthefollowingschedulingalgorithmsmustbenonpreemptive?

AX

i

/SJF

B\

1

7RR

c\

l

zFCFS

D\

l

7priorityalgorithms

A

C

Feedback:5.3.1

Difficulty:Medium

6A.Whichofthefollowingistrueofmultilevelqueuescheduling?

\

J

BZProcessescanmovebetweenqueues.

x

)

czEachqueuehasitsownschedulingalgorithm.

DAqueuecannothaveabsolutepriorityoverlower-priorityqueues.

\

J

,ItisthemostgeneralCPU-schedulingalgorithm.

Ans:B

Feedback:5.3.5

Difficulty:Medium

7.ThedefaultschedulingclassforaprocessinSolarisis.

A)timesharing

B)system

C)interactive

D)real-time

Ans:A

Feedback:5.7.3

Difficulty:Easy

8.WhichofthefollowingstatementsarefalsewithregardstotheLinuxCFSscheduler?

A)EachtaskisassignedaproportionofCPUprocessingtime.

B)Lowernumericvaluesindicatehigherrelativepriorities.

C)Thereisasingle,system-widevalueofvruntime.

D)Theschedulerdoesn*tdirectlyassignpriorities.

Ans:C

Feedback:5.7.1

Difficulty:Easy

9.TheLinuxCFSscheduleridentifiesastheintervaloftimeduringwhich

everyrunnabletaskshouldrunatleastonce.

A)virtualruntime

B)targetedlatency

C)nicevalue

D)loadbalancing

Ans:B

Feedback:5.7.1

Difficulty:Medium

t

-

lAosInLittle'sformula,X,representsthe.

x

l

Bzaveragewaitingtimeinthequeue

\

J

cZaveragearrivalratefornewprocessesinthequeue

X

I

D/averagequeuelength

X

I

/averageCPUutilization

A

n

B

Feedback:5.7.2

Difficulty:Medium

11.9InSolans,whatisthetimequantum(inmilliseconds)ofaninteractivethreadwithpriority

-

3A5\5

l

z2

B\4

)

z5

cX0

)8

/

D\35

l

7

Ans:C

Section:5.7.3

Difficulty:Easy

12.InSolaris,ifaninteractivethreadwithpriority15usesitsentiretimequantum,whatisits

priorityrecalculatedto?

A)51

B)5

C)160

D)Itremainsat15

Ans:B

Feedback:5.7.3

Difficulty:Easy

13.InSolaris,ifaninteractivethreadwithpriority25iswaitingforI/O,whatisitspriority

recalculatedtowhenitiseligibletorunagain?

A)15

B)120

C)52

D)Itremainsat25

Ans:C

Feedback:5.7.3

Difficulty:Easy

14.allowsathreadtorunononlyoneprocessor.

A)Processoraffinity

B)Processorset

C)NUMA

D)Loadbalancing

Ans:A

Feedback:5.5.2

Difficulty:Medium

15.WhatisthenumericpriorityofaWindowsthreadintheNORMAL_PRIORITY_CLASS

withHIGHESTrelativepriority?

A)24

B)10

C)8

D)13

Ans:B

Feedback:5.7,2

Difficulty:Easy

16.WhatisthenumericpriorityofaWindowsthreadintheHIGH_PRIORITY_CLASSwith

ABOVE_NORMALrelativepriority?

A)24

B)1()

C)8

D)14

Ans:D

Feedback:5.7.2

Difficulty:Easy

17.WhatisthenumericpriorityofaWindowsthreadinthe

BELOW_NORMAL_PRIORITY_CLASSwithNORMALrelativepriority?

A)6

B)7

C)5

D)8

Ans:A

Feedback:5.7.2

Difficulty:Easy

18.involvesthedecisionofwhichkernelthreadtoscheduleontowhichCPU.

A)Process-contentionscope

B)System-contentionscope

C)Dispatcher

D)Round-robinscheduling

Ans:B

Feedback:5.4.1

Difficulty:Easy

19.Withathreadexecutesonaprocessoruntilalong-latencyevent(i.e.amemory

stall)occurs.

A)coarse-grainedmultithreading

B)fine-grainedmultithreading

C)virtualization

D)multicoreprocessors

Ans:A

Feedback:5.5.4

Difficulty:Medium

20.Asignificantproblemwithpriorityschedulingalgorithmsis.

A)complexity

B)starvation

C)determiningthelengthofthenextCPUburst

D)determiningthelengthofthetimequantum

Ans:B

Feedback:5.3.3

Difficulty:Medium

21.Theoccursinfirst-come-first-servedschedulingwhenaprocesswithalongCPU

burstoccupiestheCPU.

A)dispatchlatency

B)waitingtime

C)convoyeffect

D)system-contentionscope

Ans:C

Feedback:5.3.1

Difficulty:Medium

22.Therateofaperiodictaskinahardreal-timesystemis,wherepisaperiodandtis

thetime.

A)

B0)lip

p/t

D)\/t

An

Pi

A

Section:5.6.2

Difficulty:Medium

23AXWhichotthefollowingistrueoftherate-monotonicschedulingalgorithm?

/I

B\Thetaskwiththeshortestperiodwillhavethelowestpriority.

J/

c\Itusesadynamicprioritypolicy.

—/

D\CPUutilizationisboundedwhenusingthisalgorithm.

—/Itisnon-preemptive.

An

C

Section:5.6.3

Difficulty:Difficult

24.Whichofthefollowingistrueofearliest-deadline-first(EDF)schedulingalgorithm?

A)Whenaprocessbecomesrunnable,itmustannounceitsdeadlinerequirementstothe

system.

B)Deadlinesareassignedasfollowing:theearlierthedeadline,thelowerthepriority;thelater

thedeadline,thehigherthepriority.

C)Prioritiesarefixed;thatis,theycannotbeadjustedwhenanewprocessstartsrunning.

D)Itassignsprioritiesstaticallyaccordingtodeadline.

Ans:A

Section:5.6.4

Difficulty:Medium

25.Thetwogeneralapproachestoloadbalancingareand.

A)softaffinity,hardaffinity

B)coarsegrained,finegrained

C)softreal-time,hardreal-time

D)pushmigration,pullmigration

Ans:D

Section:5.5.3

Difficulty:Medium

Essay

26.Distinguishbetweencoarse-grainedandfine-grainedmultithreading.

Ans:Therearetwoapproachestomultithreadaprocessor.(1)Coarse-grainedmultithreading

allowsathreadtorunonaprocessoruntilalong-latencyevent,suchaswaitingformemory,to

occur.Whenalong-latencyeventdoesoccur,theprocessorswitchestoanotherthread.(2)

Fine-grainedmultithreadingswitchesbetweenthreadsatamuchfiner-granularity,suchas

betweeninstructions.

Feedback:5.5.4

Difficulty:Medium

27.ExplaintheconceptofaCPU-1/0burstcycle.

Ans:Thelifecycleofaprocesscanbeconsideredtoconsistofanumberofburstsbelongingto

twodifferentstates.AllprocessesconsistofCPUcyclesandI/Ooperations.Therefore,aprocess

canbemodeledasswitchingbetweenburstsofCPUexecutionandI/Owait.

Feedback:5.1.1

Difficulty:Medium

28.WhatroledoesthedispatcherplayinCPUscheduling?

Ans:ThedispatchergivescontroloftheCPUtotheprocessselectedbytheshort-term

scheduler.Toperformthistask,acontextswitch,aswitchtousermode,andajumptotheproper

locationintheuserprogramareallrequired.Thedispatchshouldbemadeasfastaspossible.

Thetimelosttothedispatcheristermeddispatchlatency.

Feedback:5.1.4

Difficulty:Medium

29.Explainthedifferencebetweenresponsetimeandturnaroundtime.Thesetimesareboth

usedtomeasuretheeffectivenessofschedulingschemes.

Ans:Turnaroundtimeisthesumoftheperiodsthataprocessisspentwaitingtogetinto

memory,waitinginthereadyqueue,executingontheCPU,anddoingI/O.Turnaroundtime

essentiallymeasurestheamountoftimeittakestoexecuteaprocess.Responsetime,onthe

otherhand,isameasureofthetimethatelapsesbetweenarequestandthefirstresponse

produced.

Feedback:5.2

Difficulty:Medium

30.WhateffectdoesthesizeofthetimequantumhaveontheperformanceofanRR

algorithm?

Ans:Atoneextreme,ifthetimequantumisextremelylarge,theRRpolicyisthesameasthe

FCFSpolicy.Ifthetimequantumisextremelysmall,theRRapproachiscalledprocessor

sharingandcreatestheappearancethateachofnprocesseshasitsownprocessorrunningat\!n

thespeedoftherealprocessor.

Feedback:5.3.4

Difficulty:Medium

31.Explaintheprocessofstarvationandhowagingcanbeusedtopreventit.

Ans:Starvationoccurswhenaprocessisreadytorunbutisstuckwaitingindefinitelyforthe

CPU.Thiscanbecaused,forexample,whenhigher-priorityprocessespreventlow-priority

processesfromevergettingtheCPU.Aginginvolvesgraduallyincreasingthepriorityofa

processsothataprocesswilleventuallyachieveahighenoughprioritytoexecuteifitwaitedfor

alongenoughperiodoftime.

Feedback:5.3.3

Difficulty:Difficult

32.Explainthefundamentaldifferencebetweenasymmetricandsymmetricmultiprocessing.

Ans:Inasymmetricmultiprocessing,allschedulingdecisions,I/O,andothersystemactivities

arehandledbyasingleprocessor,whereasinSMP,eachprocessorisself-scheduling.

Feedback:5.5.1

Difficulty:Medium

33.Describetwogeneralapproachestoloadbalancing.

Ans:Withpushmigration,aspecifictaskperiodicallycheckstheloadoneachprocessor

and一ifitfindsanimbalance-evenlydistributestheloadbymovingprocessesfrom

overloadedtoidleorless-busyprocessors.Pullmigrationoccurswhenanidleprocessorpullsa

waitingtaskfromabusyprocessor.Pushandpullmigrationareoftenimplementedinparallelon

load-balancingsystems.

Feedback:5.5.3

Difficulty:Medium

34.InWindows,howdoesthedispatcherdeterminetheorderofthreadexecution?

Ans:Thedispatcherusesa32-levelpriorityschemetodeterminetheexecutionorder.Priorities

aredividedintotwoclasses.Thevariableclasscontainsthreadshavingprioritiesfrom1to15,

andthereal-timeclasscontainsthreadshavingprioritiesfrom16to31.Thedispatcherusesa

queueforeachschedulingpriority,andtraversesthesetofqueuesfromhighesttolowestuntilit

findsathreadthatisreadytorun.Thedispatcherexecutesanidlethreadifnoreadythreadis

found.

Feedback:5.7.2

Difficulty:Difficult

35.Whatisdeterministicmodelingandwhenisitusefulinevaluatinganalgorithm?

Ans:Deterministicmodelingtakesaparticularpredeterminedworkloadanddefinesthe

performanceofeachalgorithmforthatworkload.Deterministicmodelingissimple,fast,and

givesexactnumbersforcomparisonofalgorithms.However,itrequiresexactnumbersforinput,

anditsanswersapplyonlyinthosecases.Themainusesofdeterministicmodelingare

describingschedulingalgorithmsandprovidingexamplestoindicatetrends.

Feedback:5.8,1

Difficulty:Medium

36.Whatarethetwotypesoflatencythataffecttheperformanceofreal-timesystems?

Ans:InterruptlatencyreferstotheperiodoftimefromthearrivalofaninterruptattheCPUto

thestartoftheroutinethatservicestheinterrupt.Dispatchlatencyreferstotheamountoftime

requiredfortheschedulingdispatchertostoponeprocessandstartanother.

Section:5.6.1

Difficulty:Medium

37.WhataretheadvantagesoftheEDFschedulingalgorithmovertherate-monotonic

schedulingalgorithm?

Ans:Unliketherate-monotonicalgorithm,EDFschedulingdoesnotrequirethatprocessesbe

periodic,normustaprocessrequireaconstantamountofCPUtimeperburst.Theappealof

EDFschedulingisthatitistheoreticallyoptimal-theoretically,itcanscheduleprocessessothat

eachprocesscanmeetitsdeadlinerequirementsandCPUutilizationwillbe100percent.

Section:5.6.4

Difficulty:Medium

True/False

38.Inpreemptivescheduling,thesectionsofcodeaffectedbyinterruptsmustbeguardedfrom

simultaneoususe.

Ans:True

Feedback:5.1.3

Difficulty:Medium

39.InRRscheduling,thetimequantumshouldbesmallwithrespecttothecontext-switch

time.

Ans:False

Feedback:5.3.4

Difficulty:Medium

40.Themostcomplexschedulingalgorithmisthemultilevelfeedback-queuealgorithm.

Ans:True

Feedback:5.3.6

Difficulty:Medium

41.Loadbalancingistypicallyonlynecessaryonsystemswithacommonrunqueue.

Ans:False

Feedback:5.5.3

Difficulty:Medium

42.Systemsusingaone-to-onemodel(suchasWindows,Solaris,andLinux)schedulethreads

usingprocess-contentionscope(PCS).

Ans:False

Feedback:5.4.1

Difficulty:Easy

43.SolarisandWindowsassignhigher-prioritythreads/taskslongertimequantumsand

lower-prioritytasksshortertimequantums.

Ans:False

Feedback:5.7

Difficulty:Medium

44.ASolarisinteractivethreadwithpriority15hasahigherrelativeprioritythananinteractive

threadwithpriority20

Ans:False

Feedback:5.7,3

Difficulty:Easy

45.ASolarisinteractivethreadwithatimequantumof80hasahigherprioritythanan

interactivethreadwithatimequantumof120.

Ans:True

Feedback:5.7.3

Difficulty:Easy

46.SMPsystemsthatusemulticoreprocessorstypicallyrunfasterthanSMPsystemsthatplace

eachprocessoronseparatecores.

Ans:True

Feedback:5.5