2002_05_28_errata

Erratafor“USBRevision2.0April27,2000”asofMay28,2002:Checkthe:ThereispotentialconfusionovertheendpointnumberusedintheINpacketsenttodeviceswhenintestmodeTest_SE0_NAK.Noknownimpacttohostcontrollersordevices.Change:p169,section7.1.20,1stbullet,3rdsentence:Changesentence,“Inaddition,whileinthismode,upstreamfacingports(andonlyupstreamfacingports)mustrespondtoanyINtokenpacketdirectedtothedevice(withanendpointnumbersupportedbythedevice),andoptionallytoanyINtokenpacketatall,withaNAKhandshake(onlyifthepacketCRCisdeterminedtobecorrect)withinthenormaldeviceresponsetime.Note:Thismeansthatthehostshouldsendtestpacketstoadeviceundertestwiththedevice’scurrentaddressandendpointnumberzerotoensurethatthedevicewillrespond.”Modifycableattenuationvaluestoallow5metercables:Background:TheUSB2.0specificationalwaysintendedtoallowcablesupto5metersinlength.Analysisoftheallowedcableattenuationvaluesinthespechasshownthatthevalueat12.0Mhzistransposedandshouldbe.76dbnot.67db.Change:p.167,table7-6:Change12.0Mhzattenuationvaluefrom.67to.76.Aportionofthecorrectedtableisshownbelow.FrequencyMhzdb.5120.13.7720.151.00.204.00.398.00.5712.00.7624.00.9548.01.3596.01.90200.03.20400.05.80Chapter8ClarifytheintendedbehaviorandspecificationofNAK’dOUTs:Background:ThereisaconfusionoverthecorrectusageandinterpretationofNAKresponsestoHigh-speedbulk/controlOUTtransactions.ThischangeandacorrespondingchangeinChapter9clarifytherequiredaspectsofthisUSBfeature.Theintentwas1)torequireadevicetospecifythemaximumnumberofOUTsitcouldNAKinsomeperiod(thiswouldbetestedaspartofcompliance)and2)toallowthehosttolimittheattemptedOUTtransactionstothenumberspecifiedbythedevice.Thecurrentspecificationlanguageisinconsistentandsomewhatconfusingindescribingthis.Thischangeimpactshigh-speeddeviceswithbulkOUTandcontrolendpoints.Change:p.217,section8.5.1,5thparagraph,after3rdsentence:Breakfornewparagraph.Changesentencestarting“IfadevicerespondswithaNAKina(micro)frame,thehost...”to“IfadevicerespondswithaNAKtoanOUTinamicroframe,thehost...”Chapter9Correctuseofversion2.0inthebcdUSBfieldofthedevicedescriptor.Background:Therehasbeenconfusionoverwhen0200shouldbeusedinthebcdUSBfieldinthedevicedescriptor.Manypeoplehaveassumedthat0200issomehowassociatedwithhighspeedperipheralsandthatfullspeedorlowspeedperipheralsshouldnotuseitfortheirbcdUSBvalue.Thisisnotcorrect.AnyperipheraldesignedtotheUSB2.0specificationshoulduse0200asthebcdUSBvalue.Change:Page261,Section9.6.1,3rdparagraph.Changeparagraphtothefollowing:“TheDEVICEdescriptorofahigh-speedcapabledevicehasaversionnumberof2.0(0200H).Fullspeedandlowspeedonlydevicesdesignedtothisspecificationshouldalsouseversionnumber2.0(0200H).Ifthedeviceisfull-speedorlow-speedonlythisversionnumberindicatesthatitwillrespondcorrectlytoarequestforthedevice_qualifierdescriptor(i.e.,itwillrespondwitharequesterror).”ClarifytheintendedbehaviorandspecificationofNAK’dOUTs:Background:ThereisaconfusionoverthecorrectusageandinterpretationofNAKresponsestoHigh-speedbulk/controlOUTtransactions.ThischangeandacorrespondingchangeinChapter9clarifytherequiredaspectsofthisUSBfeature.Theintentwas1)torequireadevicetospecifythemaximumnumberofOUTsitcouldNAKinsomeperiod(thiswouldbetestedaspartofcompliance)and2)toallowthehosttolimittheattemptedOUTtransactionstothenumberspecifiedbythedevice.Thecurrentspecificationlanguageisinconsistentandsomewhatconfusingindescribingthis.Thischangeimpactshigh-speeddeviceswithbulkOUTandcontrolendpoints.Change:p.271,table9-13,descriptionofbInterval:changebulk/controldescriptionfrom“…atmost1NAKeach…”to“…afteraNAKtoanOUTthatthenextretryisafter…”.Atendofparagraph,add“SeeSection8.5.1formorePINGdetail.”Chapter10Addhostsoftwarerequirementtosuspenddownstreamportsbeforesuspendingahub:Background:Thereisaveryunlikelypossibilityofaraceconditionbetweenaremotewakeupeventandtheprocessofsuspendingahubthatcouldcreateunintendedsignalingevents.Thisissuecanbeavoidedifsystemsoftwaresuspendsadownstreamfacinghubportsbeforesuspendingahub.Change:p.282,section10.2.7,Addnewparagraph,“TheUSBsystemshouldsuspendalldownstreamfacingportsofahubbeforesuspendingthehub.”Chapter11AlloworangeinadditiontoamberLEDsfordownstreamfacinghubports:Background:SomeLEDvendorsdonotprovideanyGreen/Amber(bi-color)LEDs.However,almostallLEDvendorsprovideGreen/OrangeLEDs.AlloweitherAmberorOrangebi-colorLEDs.Change:Changereferencesfrom“amber”to“amber(ororange)”Differentportindicatorbehaviorforhubswith/withoutpowerswitching.Background:Thespecificationcurrentlyindicatesthatportindicatorstatesarenotnecessarilyconsistentinautomaticmodeforhubswith/withoutpowerswitching.Thisspecallowanceisbeingremoved.Change:p.316,Table11-6.Removethecolumnforpowerswitchingandtherowforwithoutpowerswitching.Portindicatorbehaviorthroughahubsuspend/resumecycle.Background:Therehasbeensomeconfusionoverhowtheportindicatorsonahubshouldbehaveifthehubissuspendedandthenresumed.Thecorespecificationstatesthatportindicatorsshouldbeoffwhenahubissuspended