多核處理器LINUX的進(jìn)程綁定處理器核運(yùn)行

多核辦理器Linux的進(jìn)度綁定辦理器核運(yùn)轉(zhuǎn)名詞CPUaffinity：中文稱作“CPU親和力”，是指在CMP架構(gòu)下，能夠?qū)⒁粋€(gè)或多個(gè)進(jìn)度綁定到一個(gè)或多個(gè)辦理器上運(yùn)轉(zhuǎn)。一、Linux代碼中綁定多核運(yùn)轉(zhuǎn)1、假如自己寫代碼，要把進(jìn)度綁定到CPU，該怎么做？能夠用sched_setaffinity函數(shù)。在Linux上，這會(huì)觸發(fā)一次系統(tǒng)調(diào)用。intsched_setaffinity(pid_tpid,unsignedintlen,unsignedlong*mask);sched_setaffinity的第一個(gè)參數(shù)是pid（進(jìn)度ID），設(shè)置進(jìn)度為pid的這個(gè)進(jìn)度，讓它運(yùn)轉(zhuǎn)在mask所設(shè)定的CPU上。假如pid的值為0，則表示指定的是目行進(jìn)度，使目行進(jìn)度運(yùn)轉(zhuǎn)在mask所設(shè)定的那些CPU上；第二個(gè)參數(shù)cpusetsize是mask所指定的數(shù)的長度。往常設(shè)定為sizeof(cpu_set_t)；假如目前pid所指定的CPU此時(shí)沒有運(yùn)轉(zhuǎn)在mask所指定的隨意一個(gè)CPU上,則該指定的進(jìn)度會(huì)從其余CPU上遷徙到mask的指定的一個(gè)CPU上運(yùn)轉(zhuǎn)。intsched_getaffinity(pid_tpid,unsignedintlen,unsignedlong*mask);該函數(shù)獲取pid所指示的進(jìn)度的CPU位掩碼，并將該掩碼返回到mask所指向的構(gòu)造中，即獲取指定pid目前能夠運(yùn)轉(zhuǎn)在哪些CPU上。相同，假如pid的值為0.也表示的是目行進(jìn)度。voidCPU_ZERO(cpu_set_t*set)這個(gè)宏對(duì)CPU集set進(jìn)行初始化，將其設(shè)置為空集。voidCPU_SET(intcpu,cpu_set_t*set)這個(gè)宏將cpu加入CPU集set中。voidCPU_CLR(intcpu,cpu_set_t*set)這個(gè)宏將cpu從CPU集set中刪除。intCPU_ISSET(intcpu,constcpu_set_t*set)假如cpu是CPU集set的一員，這個(gè)宏就返回一個(gè)非零值（true），不然就返回零false）。Example：我是在一個(gè)虛構(gòu)機(jī)上運(yùn)轉(zhuǎn)的程序，機(jī)器CPU是雙核的，我設(shè)置虛構(gòu)機(jī)模擬四核。在linux上履行top指令看結(jié)果，點(diǎn)擊“1”查察每個(gè)CPU核的運(yùn)轉(zhuǎn)狀況。/*Shorttestprogramtotestsched_setaffinity(whichsetstheaffinityofprocessestoprocessors).Compile:gccsched_setaffinity_test.c*-osched_setaffinity_test.o-lmUsage:./sched_setaffinity_test.oOpena"top"-windowatthesametimeandseeallthework*beingdoneonCPU0firstandafterashortwaitonCPU1，2，3.Repeatwithdifferentnumberstomakesure,itisnotacoincidence.*/#include<stdio.h>#include<math.h>#include<sched.h>doublewaste_time(longn){doubleres=0;longi=0;while(i<n*200000){i++;res+=sqrt(i);}returnres;}intmain(intargc,char**argv){unsignedlongmask=1;/*二進(jìn)制1,processor0*//*bindprocesstoprocessor0*/if(sched_setaffinity(0,sizeof(mask),&mask)<0){perror("sched_setaffinity");}/*wastesometimesotheworkisvisiblewith"top"*/printf("result:%f\n",waste_time(2000));mask=2;/*二進(jìn)制10，processswitchestoprocessor1now*/if(sched_setaffinity(0,sizeof(mask),&mask)<0){perror("sched_setaffinity");}/*wastesomemoretimetoseetheprocessorswitch*/printf("result:%f\n",waste_time(2000));mask=4;/*二進(jìn)制100，processor2*//*bindprocesstoprocessor2*/if(sched_setaffinity(0,sizeof(mask),&mask)<0){perror("sched_setaffinity");}/*wastesometimesotheworkisvisiblewith"top"*/printf("result:%f\n",waste_time(2000));mask=8;/*二進(jìn)制1000，processswitchestoprocessor3now*/if(sched_setaffinity(0,sizeof(mask),&mask)<0){perror("sched_setaffinity");}/*wastesomemoretimetoseetheprocessorswitch*/printf("result:%f\n",waste_time(2000));}2、運(yùn)轉(zhuǎn)結(jié)果圖2-1：圖中顯示編譯及履行一個(gè)使用“CPU親和力”的程序。圖2-2能夠看見CPU0已在履行程序（CPU達(dá)到100.0%？有必需嗎），第一段程序的結(jié)果還未出來。圖2-3能夠看見CPU1已在履行程序，第一段程序的結(jié)果已經(jīng)顯示（圖片中顯示了上一次履行的結(jié)果，由于截圖比較慢，履行了3次^-^），CPU1履行的結(jié)果還未顯示出來。圖2-4：能夠看見CPU2已在履行程序，第二段程序（CPU1履行）的結(jié)果已經(jīng)顯示，CPU2履行的程序結(jié)果還未出來。圖2-5：:CPU3再履行程序。圖2-6：結(jié)果已經(jīng)顯示，CPU恢復(fù)到履行以前的狀態(tài)二、多進(jìn)度（在兩個(gè)核上各自運(yùn)轉(zhuǎn)一個(gè)進(jìn)度）1、代碼以下/*Shorttestprogramtotestparallel_setaffinity_test.c(whichsetstheaffinityofprocessestoprocessors).Compile:gccparallel_setaffinity_test.c-oparallel_setaffinity_test.o-lmUsage:./parallel_setaffinity_test.o*Opena"top"-windowatthesametimeandseealltheworkbeingdoneonCPU0firstandafterashortwaitonCPU1.Repeatwithdifferentnumberstomakesure,itisnotacoincidence.*/#include<stdlib.h>#include<stdio.h>#include<sys/types.h>#include<sys/sysinfo.h>#include<unistd.h>#include<string.h>#include<math.h>#include<sched.h>doublewaste_time(longn){doubleres=0;longi=0;while(i<n*200000){i++;res+=sqrt(i);}returnres;}intmain(){charbuf[100];pid_tcld_pid;intfd;intstatus;unsignedlongmask=0;strcpy(buf,"Thisisparentprocesswrite!\n");if((cld_pid=fork())==0){strcpy(buf,"Thisischildprocesswrite!\n");printf("Thisischildprocess!\n");printf("MyPID(child)is%d\n",getpid());printf("MyPID(child)is%d\n",getppid());mask=1;/*二進(jìn)制1,processor0*//*bindprocesstoprocessor0*/if(sched_setaffinity(0,sizeof(mask),&mask)<0){perror("sched_setaffinity");}/*wastesometimesotheworkisvisiblewith"top"*/printf("result:%f.\n",waste_time(2000));write(fd,buf,strlen(buf));close(fd);exit(0);}else{printf("Thisisparentprocess!\n");printf("MyPID(parent)is%d.\n",getpid());printf("MychildPIDis%d.\n",cld_pid);mask=2;/*二進(jìn)制10，processswitchestoprocessor1now*/if(sched_setaffinity(0,sizeof(mask),&mask)<0){perror("sched_setaffinity");}/*wastesomemoretimetoseetheprocessorswitch*/printf("result:%f.\n",waste_time(2000));write(fd,buf,strlen(buf));close(fd);}wait(&status);return0;}2、運(yùn)轉(zhuǎn)結(jié)果圖3-1：圖中顯示編譯及履行一個(gè)使用“CPU親和力”的多進(jìn)度程序，雙核（虛構(gòu)四核太慢）。圖3-2：，父子進(jìn)度并發(fā)履行，父進(jìn)度在CPU1上運(yùn)轉(zhuǎn)，子進(jìn)度運(yùn)轉(zhuǎn)在CPU0，能夠看出兩個(gè)辦理器目前都在運(yùn)轉(zhuǎn)程序。圖3-3：結(jié)果已經(jīng)顯示，CPU0和CPU1都恢復(fù)到履行以前的狀態(tài)附錄1、Processoraffinityisamodificationofthenativecentralqueueschedulingalgorithminasymmetricmultiprocessingoperatingsystem.Eachtask(beitprocessorthread)inthequeuehasatagindicatingitspreferred/kinprocessor.Atallocationtime,eachtaskisallocatedtoitskinprocessorinpreferencetoothers.Processoraffinitytakesadvantageofthefactthatsomeremnantsofaprocessmayremaininoneprocessor'sstate(inparticular,initscache)fromthelasttimetheprocessran.Schedulingittorunonthesameprocessorthenexttimecouldresultintheprocess'srunningmoreefficientlybyreducingperformance-degradingsituationssuchascachemisses.Apracticalexamplemightberunningmultipleinstancesofanapplicationwhichdoesnotusemultiplethreads,suchassomegraphics-renderingsoftware.Overallsystemefficiencyincreases.Schedulingalgorithmimplementationsvaryinadherencetoprocessoraffinity.Undercertaincircumstancessomeimplementationswillallowatasktochangetoanotherprocessorifthisisdeemedtobemostefficient.Thismaybethecaseiftwoprocessor-intensivetasks(A&B)haveaffinitytooneprocessorwhileanotherprocessorliesunused.ManyalgorithmswouldshifttaskBtothesecondprocessorinordertomaximizeprocessoruse.TaskBwouldthenacquireaffinitywiththesecondprocessorwhiletaskAwouldcontinuetohaveaffinitywiththeoriginalprocessor.Processoraffinitycaneffectivelyreducecacheproblemsbutitdoesnotcurbthepersistentload-balancingproblem.[1]Further,processoraffinitybecomesmorecomplicatedinsystemswithnon-uniformarchitectures.Forexample,asystemwithtwodual-corehyper-threadedCPUspresentsachallengetoaschedulingalgorithm.ThereiscompleteaffinitybetweentwovirtualCPUsimplementedonthesamecoreviahyper-threading,partialaffinitybetweentwocoresonthesamephysicalchip(asthecoressharesome,butnotall,cache),andnoaffinitybetweenseparatephysicalchips.Asotherresourcesarealsoshared,processoraffinityalonecannotbeusedasthebasisforCPUdispatching.Ifaprocesshasrecentlyrunononevirtualhyper-threadedCPUinagivencore,andthatvirtualCPUiscurrentlybusybutitspartnerisnot,cacheaffinitywouldsuggestthattheprocessshouldbedispatchedtotheidlepartner.However,thetwovirtualCPUscompeteforessentiallyallcomputing,cache,andmemoryresources.ItwouldtypicallybemoreefficientinthiscasetodispatchtheprocesstoadifferentcoreorCPUifoneisavailable.Thiswouldlikelyincurapenaltywhenprocessrepopulatesthecache,butoverallperformancewouldlikelybehigherastheprocesswouldnothavetocompeteforresourceswithintheCPU.OnLinuxtheCPUaffinityofaprocessmightbealteredwiththetaskset(1)program.[2]OnSGIsystems,dplacebindsaprocesstoasetofCPUs.[3]NetBSD5.0,FreeBSD7.2andlaterversionscanusepthread_setaffinity_npandpthread_getaffinity_np.[4]InNetBSD,thepsrsetutility[5]tosetathread'saffinitytoacertainCPUset.InFreeBSD,cpuset[6]utilityisusedtocreateCPUsetsandtoassignprocessestothesesets.OnWindowsNT,threadandprocessCPUaffinitiescanbesetseparatelybyusingSetThreadAffinityMask[7]andSetProcessAffinityMask[8]APIcallsorviatheTaskManagerinterface(forprocessaffinityonly).MacOSXexposesanaffinityAPIthatprovideshintstothekernelhowtoschedulethreadsaccordingtoaffinitysets.2、sched_setaffinity,sched_getaffinity-setandgetaprocess'sCPUaffinitymaskSYNOPSIS#include<sched.h>intsched_setaffinity(pid_tpid,unsignedintlen,unsignedlong*mask);intsched_getaffinity(pid_tpid,unsignedintlen,unsignedlong*mask);DESCRIPTIONsched_setaffinitysetstheCPUaffinitymaskoftheprocessdenotedbypid.Ifpidiszero,thenthecurrentprocessisused.Theaffinitymaskisrepresentedbythebitmaskstoredinmask.Theleastsignificantbitcorrespondstothefirstlogicalprocessornumberonthesystem,whilethemostsignificantbitcorrespondstothelastlogicalprocessornumberonthesystem.AsetbitcorrespondstoalegallyschedulableCPUwhileanunsetbitcorrespondstoanillegallyschedulableCPU.Inotherwords,aprocessisboundtoandwillonlyrunonprocessorswhosecorrespondingbitisset.Usually,allbitsinthemaskareset.Theargumentlenisthelengthofthedatapointedtobymask.Normallythisisthesizeofawordonthesystem.ForcompatibilitywithfutureversionsoftheLinuxkernel,sincethissizecanchange,thebitmasksuppliedmustbeatleastaslargeastheaffinitymaskstoredinthekernel.Thefunctionsched_getaffinitywritesintothepointersuppliedbymaskthatissizelentheaffinitymaskofprocesspid.Ifpidiszero,thenthemaskofthecurrentprocessisreturned.RETURNVALUEOnsuccess,sched_setaffinityandsched_getaffinitybothreturn0.Onerror,-1isreturned,anderrnoissetappropriately.ERRORSEFAULTAsuppliedmemoryaddresswasinvalid.ESRCHTheprocesswhoseIDispidcouldnotbefound.EPERMThecallingprocessdoesnothaveappropriateprivileges.Theprocesscallingsched_setaffinityneedsaneffectiveuidequaltotheeuidoruidoftheprocessidentifiedbypid,oritmustpossesstheCAP_SYS_NICEcapability.EINVALTheaffinitybitmaskmaskcontainsnoprocessorsthatarephysicallyonthesystemorthelengthlenissmallerthanthesizeoftheaffinitymaskusedbythekernel.HISTORYTheaffinitysyscallswereintroducedinLinuxkernel2.5.8.Thelibrarycallswereintroducedinglibc2.3,andarestillinglibc2.3.2.Laterglibc2.3.2developmentversionschangedthisinterfacetoonewithoutthelenfield,andstilllaterversionsrevertedagain.Theglibcprototypeisnow/*SettheCPUaffinityforatask*/externintsched_setaffinity(pid_tpid,size_tcpusetsize,constcpu_set_t*cpuset);/*GettheCPUaffinityforatask*/externintsched_getaffinity(pid_tpid,3、TASKSET(1)

size_tcpusetsize,cpu_set_t*cpuset);LinuxUser’sManualTASKSET(1)NAMEtaskset-retrieveorsetaprocesses’sCPUaffinitySYNOPSIStaskset[options][mask|list][pid|command[arg]...]DESCRIPTIONtasksetisusedtosetorretrievetheCPUaffinityofarunningpro-cessgivenitsPIDortolaunchanewCOMMANDwithagivenCPUaffin-ity.CPUaffinityisaschedulerpropertythat"bonds"aprocesstoagivensetofCPUsonthesystem.TheLinuxschedulerwillhonorthegivenCPUaffinityandtheprocesswillnotrunonanyotherCPUs.NotethattheLinuxscheduleralsosupportsnaturalCPUaffinity:theschedulerattemptstokeepprocessesonthesameCPUaslongaspracti-calforperformancereasons.Therefore,forcingaspecificCPUaffin-ityisusefulonlyincertainapplications.TheCPUaffinityisrepresentedasabitmask,withthelowestorderbitcorrespondingtothefirstlogicalCPUandthehighestorderbitcorre-spondingtothelastlogicalCPU.NotallCPUsmayexistonagivensystembutamaskmayspecifymoreCPUsthanarepresent.AretrievedmaskwillreflectonlythebitsthatcorrespondtoCPUsphysicallyonthesystem.Ifaninvalidmaskisgiven(i.e.,onethatcorrespondstonovalidCPUsonthecurrentsystem)anerrorisreturned.Themasksaretypicallygiveninhexadec