NEWS 353 KB

12345678910111213141516171819202122232425262728293031323334353637383940414243444546474849505152535455565758596061626364656667686970717273747576777879808182838485868788899091929394959697989910010110210310410510610710810911011111211311411511611711811912012112212312412512612712812913013113213313413513613713813914014114214314414514614714814915015115215315415515615715815916016116216316416516616716816917017117217317417517617717817918018118218318418518618718818919019119219319419519619719819920020120220320420520620720820921021121221321421521621721821922022122222322422522622722822923023123223323423523623723823924024124224324424524624724824925025125225325425525625725825926026126226326426526626726826927027127227327427527627727827928028128228328428528628728828929029129229329429529629729829930030130230330430530630730830931031131231331431531631731831932032132232332432532632732832933033133233333433533633733833934034134234334434534634734834935035135235335435535635735835936036136236336436536636736836937037137237337437537637737837938038138238338438538638738838939039139239339439539639739839940040140240340440540640740840941041141241341441541641741841942042142242342442542642742842943043143243343443543643743843944044144244344444544644744844945045145245345445545645745845946046146246346446546646746846947047147247347447547647747847948048148248348448548648748848949049149249349449549649749849950050150250350450550650750850951051151251351451551651751851952052152252352452552652752852953053153253353453553653753853954054154254354454554654754854955055155255355455555655755855956056156256356456556656756856957057157257357457557657757857958058158258358458558658758858959059159259359459559659759859960060160260360460560660760860961061161261361461561661761861962062162262362462562662762862963063163263363463563663763863964064164264364464564664764864965065165265365465565665765865966066166266366466566666766866967067167267367467567667767867968068168268368468568668768868969069169269369469569669769869970070170270370470570670770870971071171271371471571671771871972072172272372472572672772872973073173273373473573673773873974074174274374474574674774874975075175275375475575675775875976076176276376476576676776876977077177277377477577677777877978078178278378478578678778878979079179279379479579679779879980080180280380480580680780880981081181281381481581681781881982082182282382482582682782882983083183283383483583683783883984084184284384484584684784884985085185285385485585685785885986086186286386486586686786886987087187287387487587687787887988088188288388488588688788888989089189289389489589689789889990090190290390490590690790890991091191291391491591691791891992092192292392492592692792892993093193293393493593693793893994094194294394494594694794894995095195295395495595695795895996096196296396496596696796896997097197297397497597697797897998098198298398498598698798898999099199299399499599699799899910001001100210031004100510061007100810091010101110121013101410151016101710181019102010211022102310241025102610271028102910301031103210331034103510361037103810391040104110421043104410451046104710481049105010511052105310541055105610571058105910601061106210631064106510661067106810691070107110721073107410751076107710781079108010811082108310841085108610871088108910901091109210931094109510961097109810991100110111021103110411051106110711081109111011111112111311141115111611171118111911201121112211231124112511261127112811291130113111321133113411351136113711381139114011411142114311441145114611471148114911501151115211531154115511561157115811591160116111621163116411651166116711681169117011711172117311741175117611771178117911801181118211831184118511861187118811891190119111921193119411951196119711981199120012011202120312041205120612071208120912101211121212131214121512161217121812191220122112221223122412251226122712281229123012311232123312341235123612371238123912401241124212431244124512461247124812491250125112521253125412551256125712581259126012611262126312641265126612671268126912701271127212731274127512761277127812791280128112821283128412851286128712881289129012911292129312941295129612971298129913001301130213031304130513061307130813091310131113121313131413151316131713181319132013211322132313241325132613271328132913301331133213331334133513361337133813391340134113421343134413451346134713481349135013511352135313541355135613571358135913601361136213631364136513661367136813691370137113721373137413751376137713781379138013811382138313841385138613871388138913901391139213931394139513961397139813991400140114021403140414051406140714081409141014111412141314141415141614171418141914201421142214231424142514261427142814291430143114321433143414351436143714381439144014411442144314441445144614471448144914501451145214531454145514561457145814591460146114621463146414651466146714681469147014711472147314741475147614771478147914801481148214831484148514861487148814891490149114921493149414951496149714981499150015011502150315041505150615071508150915101511151215131514151515161517151815191520152115221523152415251526152715281529153015311532153315341535153615371538153915401541154215431544154515461547154815491550155115521553155415551556155715581559156015611562156315641565156615671568156915701571157215731574157515761577157815791580158115821583158415851586158715881589159015911592159315941595159615971598159916001601160216031604160516061607160816091610161116121613161416151616161716181619162016211622162316241625162616271628162916301631163216331634163516361637163816391640164116421643164416451646164716481649165016511652165316541655165616571658165916601661166216631664166516661667166816691670167116721673167416751676167716781679168016811682168316841685168616871688168916901691169216931694169516961697169816991700170117021703170417051706170717081709171017111712171317141715171617171718171917201721172217231724172517261727172817291730173117321733173417351736173717381739174017411742174317441745174617471748174917501751175217531754175517561757175817591760176117621763176417651766176717681769177017711772177317741775177617771778177917801781178217831784178517861787178817891790179117921793179417951796179717981799180018011802180318041805180618071808180918101811181218131814181518161817181818191820182118221823182418251826182718281829183018311832183318341835183618371838183918401841184218431844184518461847184818491850185118521853185418551856185718581859186018611862186318641865186618671868186918701871187218731874187518761877187818791880188118821883188418851886188718881889189018911892189318941895189618971898189919001901190219031904190519061907190819091910191119121913191419151916191719181919192019211922192319241925192619271928192919301931193219331934193519361937193819391940194119421943194419451946194719481949195019511952195319541955195619571958195919601961196219631964196519661967196819691970197119721973197419751976197719781979198019811982198319841985198619871988198919901991199219931994199519961997199819992000200120022003200420052006200720082009201020112012201320142015201620172018201920202021202220232024202520262027202820292030203120322033203420352036203720382039204020412042204320442045204620472048204920502051205220532054205520562057205820592060206120622063206420652066206720682069207020712072207320742075207620772078207920802081208220832084208520862087208820892090209120922093209420952096209720982099210021012102210321042105210621072108210921102111211221132114211521162117211821192120212121222123212421252126212721282129213021312132213321342135213621372138213921402141214221432144214521462147214821492150215121522153215421552156215721582159216021612162216321642165216621672168216921702171217221732174217521762177217821792180218121822183218421852186218721882189219021912192219321942195219621972198219922002201220222032204220522062207220822092210221122122213221422152216221722182219222022212222222322242225222622272228222922302231223222332234223522362237223822392240224122422243224422452246224722482249225022512252225322542255225622572258225922602261226222632264226522662267226822692270227122722273227422752276227722782279228022812282228322842285228622872288228922902291229222932294229522962297229822992300230123022303230423052306230723082309231023112312231323142315231623172318231923202321232223232324232523262327232823292330233123322333233423352336233723382339234023412342234323442345234623472348234923502351235223532354235523562357235823592360236123622363236423652366236723682369237023712372237323742375237623772378237923802381238223832384238523862387238823892390239123922393239423952396239723982399240024012402240324042405240624072408240924102411241224132414241524162417241824192420242124222423242424252426242724282429243024312432243324342435243624372438243924402441244224432444244524462447244824492450245124522453245424552456245724582459246024612462246324642465246624672468246924702471247224732474247524762477247824792480248124822483248424852486248724882489249024912492249324942495249624972498249925002501250225032504250525062507250825092510251125122513251425152516251725182519252025212522252325242525252625272528252925302531253225332534253525362537253825392540254125422543254425452546254725482549255025512552255325542555255625572558255925602561256225632564256525662567256825692570257125722573257425752576257725782579258025812582258325842585258625872588258925902591259225932594259525962597259825992600260126022603260426052606260726082609261026112612261326142615261626172618261926202621262226232624262526262627262826292630263126322633263426352636263726382639264026412642264326442645264626472648264926502651265226532654265526562657265826592660266126622663266426652666266726682669267026712672267326742675267626772678267926802681268226832684268526862687268826892690269126922693269426952696269726982699270027012702270327042705270627072708270927102711271227132714271527162717271827192720272127222723272427252726272727282729273027312732273327342735273627372738273927402741274227432744274527462747274827492750275127522753275427552756275727582759276027612762276327642765276627672768276927702771277227732774277527762777277827792780278127822783278427852786278727882789279027912792279327942795279627972798279928002801280228032804280528062807280828092810281128122813281428152816281728182819282028212822282328242825282628272828282928302831283228332834283528362837283828392840284128422843284428452846284728482849285028512852285328542855285628572858285928602861286228632864286528662867286828692870287128722873287428752876287728782879288028812882288328842885288628872888288928902891289228932894289528962897289828992900290129022903290429052906290729082909291029112912291329142915291629172918291929202921292229232924292529262927292829292930293129322933293429352936293729382939294029412942294329442945294629472948294929502951295229532954295529562957295829592960296129622963296429652966296729682969297029712972297329742975297629772978297929802981298229832984298529862987298829892990299129922993299429952996299729982999300030013002300330043005300630073008300930103011301230133014301530163017301830193020302130223023302430253026302730283029303030313032303330343035303630373038303930403041304230433044304530463047304830493050305130523053305430553056305730583059306030613062306330643065306630673068306930703071307230733074307530763077307830793080308130823083308430853086308730883089309030913092309330943095309630973098309931003101310231033104310531063107310831093110311131123113311431153116311731183119312031213122312331243125312631273128312931303131313231333134313531363137313831393140314131423143314431453146314731483149315031513152315331543155315631573158315931603161316231633164316531663167316831693170317131723173317431753176317731783179318031813182318331843185318631873188318931903191319231933194319531963197319831993200320132023203320432053206320732083209321032113212321332143215321632173218321932203221322232233224322532263227322832293230323132323233323432353236323732383239324032413242324332443245324632473248324932503251325232533254325532563257325832593260326132623263326432653266326732683269327032713272327332743275327632773278327932803281328232833284328532863287328832893290329132923293329432953296329732983299330033013302330333043305330633073308330933103311331233133314331533163317331833193320332133223323332433253326332733283329333033313332333333343335333633373338333933403341334233433344334533463347334833493350335133523353335433553356335733583359336033613362336333643365336633673368336933703371337233733374337533763377337833793380338133823383338433853386338733883389339033913392339333943395339633973398339934003401340234033404340534063407340834093410341134123413341434153416341734183419342034213422342334243425342634273428342934303431343234333434343534363437343834393440344134423443344434453446344734483449345034513452345334543455345634573458345934603461346234633464346534663467346834693470347134723473347434753476347734783479348034813482348334843485348634873488348934903491349234933494349534963497349834993500350135023503350435053506350735083509351035113512351335143515351635173518351935203521352235233524352535263527352835293530353135323533353435353536353735383539354035413542354335443545354635473548354935503551355235533554355535563557355835593560356135623563356435653566356735683569357035713572357335743575357635773578357935803581358235833584358535863587358835893590359135923593359435953596359735983599360036013602360336043605360636073608360936103611361236133614361536163617361836193620362136223623362436253626362736283629363036313632363336343635363636373638363936403641364236433644364536463647364836493650365136523653365436553656365736583659366036613662366336643665366636673668366936703671367236733674367536763677367836793680368136823683368436853686368736883689369036913692369336943695369636973698369937003701370237033704370537063707370837093710371137123713371437153716371737183719372037213722372337243725372637273728372937303731373237333734373537363737373837393740374137423743374437453746374737483749375037513752375337543755375637573758375937603761376237633764376537663767376837693770377137723773377437753776377737783779378037813782378337843785378637873788378937903791379237933794379537963797379837993800380138023803380438053806380738083809381038113812381338143815381638173818381938203821382238233824382538263827382838293830383138323833383438353836383738383839384038413842384338443845384638473848384938503851385238533854385538563857385838593860386138623863386438653866386738683869387038713872387338743875387638773878387938803881388238833884388538863887388838893890389138923893389438953896389738983899390039013902390339043905390639073908390939103911391239133914391539163917391839193920392139223923392439253926392739283929393039313932393339343935393639373938393939403941394239433944394539463947394839493950395139523953395439553956395739583959396039613962396339643965396639673968396939703971397239733974397539763977397839793980398139823983398439853986398739883989399039913992399339943995399639973998399940004001400240034004400540064007400840094010401140124013401440154016401740184019402040214022402340244025402640274028402940304031403240334034403540364037403840394040404140424043404440454046404740484049405040514052405340544055405640574058405940604061406240634064406540664067406840694070407140724073407440754076407740784079408040814082408340844085408640874088408940904091409240934094409540964097409840994100410141024103410441054106410741084109411041114112411341144115411641174118411941204121412241234124412541264127412841294130413141324133413441354136413741384139414041414142414341444145414641474148414941504151415241534154415541564157415841594160416141624163416441654166416741684169417041714172417341744175417641774178417941804181418241834184418541864187418841894190419141924193419441954196419741984199420042014202420342044205420642074208420942104211421242134214421542164217421842194220422142224223422442254226422742284229423042314232423342344235423642374238423942404241424242434244424542464247424842494250425142524253425442554256425742584259426042614262426342644265426642674268426942704271427242734274427542764277427842794280428142824283428442854286428742884289429042914292429342944295429642974298429943004301430243034304430543064307430843094310431143124313431443154316431743184319432043214322432343244325432643274328432943304331433243334334433543364337433843394340434143424343434443454346434743484349435043514352435343544355435643574358435943604361436243634364436543664367436843694370437143724373437443754376437743784379438043814382438343844385438643874388438943904391439243934394439543964397439843994400440144024403440444054406440744084409441044114412441344144415441644174418441944204421442244234424442544264427442844294430443144324433443444354436443744384439444044414442444344444445444644474448444944504451445244534454445544564457445844594460446144624463446444654466446744684469447044714472447344744475447644774478447944804481448244834484448544864487448844894490449144924493449444954496449744984499450045014502450345044505450645074508450945104511451245134514451545164517451845194520452145224523452445254526452745284529453045314532453345344535453645374538453945404541454245434544454545464547454845494550455145524553455445554556455745584559456045614562456345644565456645674568456945704571457245734574457545764577457845794580458145824583458445854586458745884589459045914592459345944595459645974598459946004601460246034604460546064607460846094610461146124613461446154616461746184619462046214622462346244625462646274628462946304631463246334634463546364637463846394640464146424643464446454646464746484649465046514652465346544655465646574658465946604661466246634664466546664667466846694670467146724673467446754676467746784679468046814682468346844685468646874688468946904691469246934694469546964697469846994700470147024703470447054706470747084709471047114712471347144715471647174718471947204721472247234724472547264727472847294730473147324733473447354736473747384739474047414742474347444745474647474748474947504751475247534754475547564757475847594760476147624763476447654766476747684769477047714772477347744775477647774778477947804781478247834784478547864787478847894790479147924793479447954796479747984799480048014802480348044805480648074808480948104811481248134814481548164817481848194820482148224823482448254826482748284829483048314832483348344835483648374838483948404841484248434844484548464847484848494850485148524853485448554856485748584859486048614862486348644865486648674868486948704871487248734874487548764877487848794880488148824883488448854886488748884889489048914892489348944895489648974898489949004901490249034904490549064907490849094910491149124913491449154916491749184919492049214922492349244925492649274928492949304931493249334934493549364937493849394940494149424943494449454946494749484949495049514952495349544955495649574958495949604961496249634964496549664967496849694970497149724973497449754976497749784979498049814982498349844985498649874988498949904991499249934994499549964997499849995000500150025003500450055006500750085009501050115012501350145015501650175018501950205021502250235024502550265027502850295030503150325033503450355036503750385039504050415042504350445045504650475048504950505051505250535054505550565057505850595060506150625063506450655066506750685069507050715072507350745075507650775078507950805081508250835084508550865087508850895090509150925093509450955096509750985099510051015102510351045105510651075108510951105111511251135114511551165117511851195120512151225123512451255126512751285129513051315132513351345135513651375138513951405141514251435144514551465147514851495150515151525153515451555156515751585159516051615162516351645165516651675168516951705171517251735174517551765177517851795180518151825183518451855186518751885189519051915192519351945195519651975198519952005201520252035204520552065207520852095210521152125213521452155216521752185219522052215222522352245225522652275228522952305231523252335234523552365237523852395240524152425243524452455246524752485249525052515252525352545255525652575258525952605261526252635264526552665267526852695270527152725273527452755276527752785279528052815282528352845285528652875288528952905291529252935294529552965297529852995300530153025303530453055306530753085309531053115312531353145315531653175318531953205321532253235324532553265327532853295330533153325333533453355336533753385339534053415342534353445345534653475348534953505351535253535354535553565357535853595360536153625363536453655366536753685369537053715372537353745375537653775378537953805381538253835384538553865387538853895390539153925393539453955396539753985399540054015402540354045405540654075408540954105411541254135414541554165417541854195420542154225423542454255426542754285429543054315432543354345435543654375438543954405441544254435444544554465447544854495450545154525453545454555456545754585459546054615462546354645465546654675468546954705471547254735474547554765477547854795480548154825483548454855486548754885489549054915492549354945495549654975498549955005501550255035504550555065507550855095510551155125513551455155516551755185519552055215522552355245525552655275528552955305531553255335534553555365537553855395540554155425543554455455546554755485549555055515552555355545555555655575558555955605561556255635564556555665567556855695570557155725573557455755576557755785579558055815582558355845585558655875588558955905591559255935594559555965597559855995600560156025603560456055606560756085609561056115612561356145615561656175618561956205621562256235624562556265627562856295630563156325633563456355636563756385639564056415642564356445645564656475648564956505651565256535654565556565657565856595660566156625663566456655666566756685669567056715672567356745675567656775678567956805681568256835684568556865687568856895690569156925693569456955696569756985699570057015702570357045705570657075708570957105711571257135714571557165717571857195720572157225723572457255726572757285729573057315732573357345735573657375738573957405741574257435744574557465747574857495750575157525753575457555756575757585759576057615762576357645765576657675768576957705771577257735774577557765777577857795780578157825783578457855786578757885789579057915792579357945795579657975798579958005801580258035804580558065807580858095810581158125813581458155816581758185819582058215822582358245825582658275828582958305831583258335834583558365837583858395840584158425843584458455846584758485849585058515852585358545855585658575858585958605861586258635864586558665867586858695870587158725873587458755876587758785879588058815882588358845885588658875888588958905891589258935894589558965897589858995900590159025903590459055906590759085909591059115912591359145915591659175918591959205921592259235924592559265927592859295930593159325933593459355936593759385939594059415942594359445945594659475948594959505951595259535954595559565957595859595960596159625963596459655966596759685969597059715972597359745975597659775978597959805981598259835984598559865987598859895990599159925993599459955996599759985999600060016002600360046005600660076008600960106011601260136014601560166017601860196020602160226023602460256026602760286029603060316032603360346035603660376038603960406041604260436044604560466047604860496050605160526053605460556056605760586059606060616062606360646065606660676068606960706071607260736074607560766077607860796080608160826083608460856086608760886089609060916092609360946095609660976098609961006101610261036104610561066107610861096110611161126113611461156116611761186119612061216122612361246125612661276128612961306131613261336134613561366137613861396140614161426143614461456146614761486149615061516152615361546155615661576158615961606161616261636164616561666167616861696170617161726173617461756176617761786179618061816182618361846185618661876188618961906191619261936194619561966197619861996200620162026203620462056206620762086209621062116212621362146215621662176218621962206221622262236224622562266227622862296230623162326233623462356236623762386239624062416242624362446245624662476248624962506251625262536254625562566257625862596260626162626263626462656266626762686269627062716272627362746275627662776278627962806281628262836284628562866287628862896290629162926293629462956296629762986299630063016302630363046305630663076308630963106311631263136314631563166317631863196320632163226323632463256326632763286329633063316332633363346335633663376338633963406341634263436344634563466347634863496350635163526353635463556356635763586359636063616362636363646365636663676368636963706371637263736374637563766377637863796380638163826383638463856386638763886389639063916392639363946395639663976398639964006401640264036404640564066407640864096410641164126413641464156416641764186419642064216422642364246425642664276428642964306431643264336434643564366437643864396440644164426443644464456446644764486449645064516452645364546455645664576458645964606461646264636464646564666467646864696470647164726473647464756476647764786479648064816482648364846485648664876488648964906491649264936494649564966497649864996500650165026503650465056506650765086509651065116512651365146515651665176518651965206521652265236524652565266527652865296530653165326533653465356536653765386539654065416542654365446545654665476548654965506551655265536554655565566557655865596560656165626563656465656566656765686569657065716572657365746575657665776578657965806581658265836584658565866587658865896590659165926593659465956596659765986599660066016602660366046605660666076608660966106611661266136614661566166617661866196620662166226623662466256626662766286629663066316632663366346635663666376638663966406641664266436644664566466647664866496650665166526653665466556656665766586659666066616662666366646665666666676668666966706671667266736674667566766677667866796680668166826683668466856686668766886689669066916692669366946695669666976698669967006701670267036704670567066707670867096710671167126713671467156716671767186719672067216722672367246725672667276728672967306731673267336734673567366737673867396740674167426743674467456746674767486749675067516752675367546755675667576758675967606761676267636764676567666767676867696770677167726773677467756776677767786779678067816782678367846785678667876788678967906791679267936794679567966797679867996800680168026803680468056806680768086809681068116812681368146815681668176818681968206821682268236824682568266827682868296830683168326833683468356836683768386839684068416842684368446845684668476848684968506851685268536854685568566857685868596860686168626863686468656866686768686869687068716872687368746875687668776878687968806881688268836884688568866887688868896890689168926893689468956896689768986899690069016902690369046905690669076908690969106911691269136914691569166917691869196920692169226923692469256926692769286929693069316932693369346935693669376938693969406941694269436944694569466947694869496950695169526953695469556956695769586959696069616962696369646965696669676968696969706971697269736974697569766977697869796980698169826983698469856986698769886989699069916992699369946995699669976998699970007001700270037004700570067007700870097010701170127013701470157016701770187019702070217022702370247025702670277028702970307031703270337034703570367037703870397040704170427043704470457046704770487049705070517052705370547055705670577058705970607061706270637064706570667067706870697070707170727073707470757076707770787079708070817082708370847085708670877088708970907091709270937094709570967097709870997100710171027103710471057106710771087109711071117112711371147115711671177118711971207121712271237124712571267127712871297130713171327133713471357136713771387139714071417142714371447145714671477148714971507151715271537154715571567157715871597160716171627163716471657166716771687169717071717172717371747175717671777178717971807181718271837184718571867187718871897190719171927193719471957196719771987199720072017202720372047205720672077208720972107211721272137214721572167217721872197220722172227223722472257226722772287229723072317232723372347235723672377238723972407241724272437244724572467247724872497250725172527253725472557256725772587259726072617262726372647265726672677268726972707271727272737274727572767277727872797280728172827283728472857286728772887289729072917292729372947295729672977298729973007301730273037304730573067307730873097310731173127313731473157316731773187319732073217322732373247325732673277328732973307331733273337334733573367337733873397340734173427343734473457346734773487349735073517352735373547355735673577358735973607361736273637364736573667367736873697370737173727373737473757376737773787379738073817382738373847385738673877388738973907391739273937394739573967397739873997400
  1. Version 4.11.1 - 16th August 2018
  2. - Alias T1factory option to noauto.
  3. - sem_timedwait needs to use CLOCK_REALTIME.
  4. - Conditional timedwaits need to use CLOCK_REALTIME.
  5. - Remove unused abstime.
  6. Version 4.11.0 - 15th August 2018
  7. - Disable implicit fallthrough warning for newer GCCs.
  8. - Alphabetical options.
  9. - Cope with abs timeouts not working on some OSs, fixing a logic error on
  10. cgsem_mswait.
  11. - hash table: use lookup instead of iteration
  12. - Do away with useless factory tune.
  13. - Keep diff low for accurate hashrates.
  14. - Take an optional fanspeed target allowing for quiet operation.
  15. - Change default pll back to 1332.
  16. - Set lastshare on successful reinit and sleep longer for each retry.
  17. - Allow much tighter temperature control if temperature is unchanged for a
  18. longer period, allowing temps to get to 75 degrees and have optimally low
  19. fanspeed.
  20. - Use clock_gettime instead of gettimeofday to not be burnt by ntp changes.
  21. - Remove useless variables.
  22. - Should be 390k.
  23. - Set lower spi speed to allow ultra low frequencies on shutdown.
  24. - Re-initialise chains first if possible instead of shutting down cgminer.
  25. - Abstract out prepare T1.
  26. - Abstract out start t1 chain.
  27. - More reliably shut down cgminer from the driver thread.
  28. - Import SPI fixes.
  29. - Import dev fixes.
  30. - Speed up mcompat_find_chain_vid since we are just going to lower voltage
  31. during tuning it doesn't need to be very accurate.
  32. - Fast start/restart with clean shutdown to lower power but powered on mode.
  33. - Import Dragonmint T1 driver and version rolling AKA asicboost support.
  34. - Do not give out API description to random probes to inappopriately open ports.
  35. - Use system provided uthash and libjansson if provided preferentially now.
  36. - Update uthash to version 2.0.2
  37. - Avoid potential buffer overflows in api.c
  38. - Fix write config with enable-icarus changes.
  39. - Sleep if btcd isn't immediately available and try again when attempting to
  40. decode a coinbase.
  41. - Elaborate what type of pool coinbase transactions to decode and do not print a
  42. warning if no devices are attached when decoding.
  43. - Add a --decode option that decodes a pool's coinbase transactions and exits.
  44. - Set up gbt_curl before setting gbt_solo flag to prevent gbt_curl deref in
  45. longpoll_thread.
  46. - remove cgminer-api libusb dependency
  47. - Update jasson library in gitignore
  48. - Fix compile guide for api-example.c
  49. - Cleanup libusb-1.0 in gitignore
  50. - Remove AVA7_FREQ_CUTOFF_MODE for Avalon7
  51. - Initialize default temp in detect_modules for Avalon7
  52. - Remove AVA7_FREQ_TEMPADJ_MODE for Avalon7
  53. - Fix build with disable-libcurl.
  54. Version 4.10.0 - 23rd January 2017
  55. - Silence warning.
  56. - Clean up compac driver.
  57. - Allocate appropriate memory size for PSECURITY_DESCRIPTOR
  58. - Remove jansson memory tricks.
  59. - Increase header size for gbt solo.
  60. - Increase header size.
  61. - Fix curl build on mingw
  62. - Windows build fixes.
  63. - Fix avalon4 warnings.
  64. - We can only change diff once per notify so assume successive diffs are stacked
  65. for successive notifies.
  66. - Enable avalon4 building with avalon7.
  67. - Move to system libusb.
  68. - Upgrade to jansson 2.9
  69. - display restart thread errno
  70. - String length sanity checks in config parsing.
  71. - Make pool_no and block height logging more consistent
  72. - Always clean swork when prev_hash changes
  73. - Add nonce mask to api for Avalon7
  74. - Add new options for Avalon7
  75. - Add a device table for Avalon7
  76. - Update AVA7_DEFAULT_MINER_CNT usage
  77. - Add frequency range support 100MHz - 500MHz
  78. - Added roll_work_ntime() to enchance performance
  79. - Revert "Set default ava7 nonce mask to 27 which empiric testing shows stable
  80. and more rapid rise in frequency without significant overshoot."
  81. - Remove useless code in avalon7
  82. - Detect avalon7 modules during ava7 init and only add the device if any modules
  83. exist.
  84. - Unplug avalon7 on failing to re-init AUC to allow a clean hotplug event which
  85. is more reliable.
  86. - Fix warning without ava4/7
  87. - Re-initialise libusb if there are no active devices before the next hotplug
  88. event.
  89. - Reset the tv_end variable before each libusb handle events call.
  90. - Fix Vo display for A741
  91. - Update AVA7_DEFAULT_ASIC_MAX for A741
  92. - Tweak ava7 fan control further.
  93. - Set default ava7 nonce mask to 27 which empiric testing shows stable and more
  94. rapid rise in frequency without significant overshoot.
  95. - Set ava7 starting fan to 1/3 speed between min and max.
  96. - Convert the avalon7 fan control to use a PID-like feedback mechanism for
  97. optimal fan speed and quasi-constant temperature. Change meaning of target
  98. temperature to actual desired temperature, setting it to 90C. Set minimum
  99. fanspeed to 5% as a safety fallback.
  100. - Inherit the diff from the original stratum in submit_nonce2_nonce, fixing
  101. wrong diff submits on avalon6/7.
  102. - Use diff1 and diff rejected to calculate hashrate in avalon7 for a more stable
  103. displayed hashrate.
  104. - Check for nodev errors after usb transfers to not change the err value.
  105. - There is no need for complicated locking in usb_reset; simply use the write
  106. lock and avoid locking risks.
  107. - Update nonce mask from 31 to 29
  108. - Send jobid when it's update
  109. - Flush useless works when new block was found
  110. - Update error polling count
  111. - Fix frequency mode for Avalon7
  112. - Added connection overload detection and API entry
  113. - Fixed some English mistakes
  114. - Fix display mm status when stratum pool is failed
  115. - Update api display
  116. - Reset usb devices if they fail to initialise.
  117. - TIMER_ABSTIME is not available on osx
  118. - Get rid of binary stratum work based on avalon7 define.
  119. - Remove impossible comparisons from avalon7.c
  120. - Style police.
  121. - Don't use the stratum work generator for Avalon7
  122. - Add STRATUM_USER_AGENT macro for utils
  123. - Update README and ASIC-README for Avalon7
  124. - Add Avalon7 support
  125. - Fix warning in AV4 driver
  126. - Update AVA4_DEFAULT_MODULARS
  127. - Update nonce2 start and range
  128. - Add more options for AVA4_FREQ_TEMPADJ_MODE
  129. - Double check data size when use avalon4_auc_xfer
  130. - Update AVA4_MM60_TEMP_FREQADJ
  131. - Fix xfer_err_cnt
  132. - Avoid thread block when use AV4 iic driver
  133. - Update check rules for AUC
  134. - Support AV4 miner detect through iic
  135. - Fix some settings for AvalonMiner 6.0
  136. - advertise segwit support for solo mining
  137. - add support for segwit commitment insertion
  138. - add bip9 support
  139. - Fix work update for hash_work
  140. - Fix various minor once-off memory leaks
  141. - Fix read beyond stack size issue
  142. - Reattach kernel driver to correct IF
  143. - Fix clobbered by ‘longjmp’ or ‘vfork’ warning
  144. - Fix unreachable code
  145. - allow for independent target freq so bad stick doesn't effect group
  146. - +1 freq hex alignment
  147. - tidy up naming to match general conventions
  148. - tidy up some unused vars
  149. - Add in ramp code. Untie compac detect from default icarus fallback detect to
  150. have a cleaner device detect path.
  151. - move over code to fit serial number on screen
  152. - changed a few spaces to tabs
  153. - Added GekkoScience Compac detection and support to the icarus driver.
  154. - Show Current Block Version under pools api
  155. - Show Current Block Height under pools api
  156. - Update avalon4 options
  157. - Update AVA4_DEFAULT_SMART_SPEED to AVA4_DEFAULT_SMARTSPEED_MODE3
  158. - Update avalon4_update process
  159. - Code cleanup
  160. - Minor update
  161. - Detach the duplicate dev when detect in xfer
  162. - Update mm count when scan hash
  163. - Check module if exits when detect
  164. - Update hashmeter for avalon devs
  165. - Fix i_5s usage
  166. - Update smart frequency for AVA4_DEFAULT_SMARTSPEED_MODE3
  167. - Add new smart speed mode (mode 3)
  168. - Update default value
  169. - Add more options for smartspeed
  170. - Don't label threads of devices that are paused as sick
  171. - Set avalon4 to cut off if thread should be paused
  172. - ASIC-README: Avalon6 will come up as AV6
  173. - Update the ASIC-README and help message of Avalon4/4.1/6
  174. - fix cgminer-api compilation on osx
  175. - Support low diffs like those on testnets
  176. - Restart being missed outside of block change
  177. - Set default frequencies for ava4/6 in the correct place if none is specified
  178. - Use a generic zero stats function for drivers that don't implement their own
  179. - The fan control for ava6 overshoots often so use a safer non-linear curve
  180. mechanism relative to the target temperature and remove inappropriate caps to
  181. opt_avalon4_overheat and opt_avalon4_temp_target
  182. - 450 would appear to be the most reliable startup speed for ava6 so revise it
  183. down further
  184. - Avoid double locking when checking if a block exists to prevent a further race
  185. between seeing it and adding it
  186. - AVA6 is more reliable starting up at the slightly slower speed of 470, making
  187. it internally start at 450 before working its way up
  188. - Set the freq array variables when setting default frequency as well
  189. - Set default frequency according to ava type if none is specified on the
  190. command line, choosing 475 as default for ava6
  191. - Use AV6 name for avalon6 in ava4 driver
  192. - Update copyright notices for ava4 driver
  193. - Fix warnings
  194. - Log blockheight when the pool changes to a new block.
  195. - Check for stratum clean message in test_work_current to not give false
  196. positive messages for missed pool notifications of block changes
  197. - Update AVA4_DEFAULT_SPEED_ERROR
  198. - Add new option for Avalon6
  199. - Update frequency process
  200. - Update AVA4_DEFAULT_TEMP_TARGET for Avalon6
  201. - Update voltage decode for Avalon6
  202. - Update temperature target and overheat
  203. - Update target temp settting
  204. - Update fan adjust
  205. - Update Vol api display
  206. - Update avalon4 option
  207. - Display stratum difficulty in pool api
  208. - Fix divide by zero
  209. - Fix nonce counter
  210. - Fix GHS display
  211. - Display total asics for Avalon6
  212. - Update max diff for Avalon4 and Avalon6
  213. - Update temperature check
  214. - Fix typo
  215. - By default we enable automatic frequency
  216. - Minor changes
  217. - Update display
  218. - Update api display
  219. - Fix GHSmm
  220. - Update api for Avalon6
  221. - Display pll infomation for Avalon6
  222. - Update frequency for Avalon6
  223. - Turn on avalon6 nonce check as default
  224. - Update convert voltage
  225. - Update voltage display
  226. - Display more status
  227. - Enable adjust frequency automatically
  228. - Update convert voltage
  229. - Update temperature display for Avalon6
  230. - Add more options for Avalon6
  231. - Update asic count for Avalon6
  232. - Update hashrate counter
  233. - Decode voltage and temperature for Avalon6
  234. - Support error code for Avalon6
  235. - Support 2 miners for Avalon6
  236. - Show ASCI status for Avalon6
  237. - Don't need config voltage for Avalon6
  238. - Update MM60 profile
  239. - Add nonce counter for Avalon6
  240. - Hide MW for Avalon6.0
  241. - Count hw for Avalon6
  242. - Increse stat buffer size
  243. - Fix job_idcmp when copy stratum
  244. - Add Avalon6 support
  245. - Apply immediately the first set_difficulty
  246. - store the next difficulty
  247. - Make pool fallback time configurable and default to 2 minutes instead of 5.
  248. Rework fallback mechanism to check pool status every 5 seconds and not miss a
  249. recovering pool.
  250. - Fix error message for pools such as p2pool that use no nonce1 in their stratum
  251. templates
  252. - Set work delay to 1 for AntS1 and AntS2
  253. - driver-bitmain default tempoverctrl off so it can be disabled
  254. - API add Work Difficulty to pools
  255. - remove some documention references to getwork
  256. - Support multiple modulars frequency setting
  257. - Decode frequency for Avalon nano 2.0
  258. - Decode adc status for Avalon nano 2.0
  259. - Display temperature in api (Avalon nano 2.0
  260. - Display asic match works for Avalon5
  261. - Support adjust frequency by asics
  262. - Display asic freq in status
  263. - Diplay moving average dh for Avalon5
  264. - Fix api display
  265. - Upgrade date/timestamp strings to millisecond accuracy
  266. - Support voltage adjustment automatically by modular
  267. - Support Avalon5
  268. - Increase AVAM_DEFAULT_ADJ_INTERVAL
  269. - Boundary check
  270. - Update AVAM_DEFAULT_VOLTAGE_MIN
  271. - Check the final freq
  272. - Support frequency adjust automatically
  273. - Add simple moving sum of the hardware errors
  274. - Add elapsed info for Avalon miner
  275. - Advise against building blockerupter driver
  276. - Move blockerupter driver to after icarus to not interfere with their detection
  277. if built in
  278. Version 4.9.2 - 12th June 2015
  279. - Only blacklist devices on iManufacturer if they don't have an iProduct
  280. - Simplify the pool work choice switching now that only local work generation is
  281. supported
  282. - Check for absense of stratum notify on stratum pools to see if they're usable
  283. - Remove getwork support
  284. - Remove test for localgen since all work will be
  285. - Deprecate failover only option in anticipation of removing getwork
  286. - Deprecate expiry option and fix it to 10 minutes, replacing all deprecated
  287. config calls in the API with a deprecated message
  288. - Deprecate scan-time which is irrelevant with ASIC speed mining
  289. - Deprecate opt_queue since all mining is now local work generation which is
  290. rapid and it serves no useful purpose to generate work that is potentially
  291. stale, increasing CPU usage to check it before using it
  292. - Change default au3 voltage to 775 to more reliably run at default frequency
  293. - Set correct timeout for AU3 in icarus_detect_one though it won't affect
  294. functioning
  295. - Blacklist LIX device which doesn't exist
  296. - Round up fail time to next highest integer on icarus
  297. - AntS3 - set default work delay
  298. - add pool block quality to api
  299. - AntS3 - redo work send/stats and add a tuning option - not yet tuned
  300. - Ant - ignore useless options in case they are passed
  301. - AntS3 - initial S3 merge basic support
  302. - Ant voltage isn't S1
  303. - API - restore IPv4 address .0 padding and update copyright dates
  304. - ants2 - separate freq from options
  305. - ants2 - make it compile :p
  306. - ants2 - add voltage
  307. - ants2 - remove libusb/udev dependency since it doesn't use USB
  308. - AntS1 ensure correct endian and correct wid
  309. - libsystemd: Notify watchdog
  310. - libsystemd: Notify run state and status line
  311. - Catch SIGABRT signal and exit gracefully
  312. - Add libsystemd configure option (default disabled)
  313. - Document --with-system-jansson configure option
  314. - SP30: Fix logging to use applog over printf calls
  315. - Allow building with system jansson
  316. - Set request ID in validateaddress JSON-RPC request.
  317. - Cope with reconnect being sent a port number as an integer or string
  318. - Added syslog mask to limit amount of output
  319. - regressed with "BAL" little-jalapenos
  320. - Add yet more whitelisting of BMA devices, removing the case insensitive match
  321. for manufacturer
  322. - Fix detection of no mining device configuration.
  323. - Remove LT_INIT from configure which breaks ltmain.sh detection for some
  324. automagic reason
  325. - Add missing hashtable_seed file
  326. - Add missing file
  327. - Speed up shutdown by decreasing usb poll time to 100ms and using no timeout on
  328. exit
  329. - Cope with attempted reuse of strings in input_pool
  330. - Manage failed URL entry at startup more gracefully, adding stratum+tcp://
  331. automatically if http:// isn't specified
  332. - Fix various symbol size change warnings
  333. - Import jansson updates from ckpool
  334. - Check for attempts to copy to/from null in cg_memcpy
  335. - Use cg_memcpy throughtout bitfury driver
  336. - Use cg_memcpy throughout cgminer.c and util.c
  337. - Use alloc helper in miner.h
  338. - Use alloc helpers in bitfury driver
  339. - Use alloc helpers in spondoolies drivers
  340. - Use alloc helpers in bflsc driver
  341. - Align_len in all alloc helper calls
  342. - Use align_len in all alloc helpers
  343. - Use alloc helpers in usbutils
  344. - Use alloc helpers in icarus driver
  345. - Use alloc helpers in avalon4 driver
  346. - Use alloc helpers in api.c
  347. - Use the alloc helpers in util.c
  348. - Convert all users of alloc to use the helpers in cgminer.c
  349. - Add helper functions to various alloc functions that automatically check for
  350. failure
  351. - Fix rare dereference error on pool stratum queue
  352. Version 4.9.1 - 3rd February 2015
  353. - Fix various unused warnings
  354. - Fix avalon4 warnings
  355. - Display notice if pool successfully negotiates stratum resume
  356. - Support auto adjust voltage individually
  357. - Don't keep retrying to connect to a pool that has been removed
  358. - Null the actual pointer used to call discard and free_work and safely handle
  359. being called with a null pointer, giving a verbose warning about the call site
  360. - Fix off by one error when running out of queued IDs in bflsc28_queue_full
  361. - Uninit BET driver when it fails to initialise any boards
  362. - Fix detection of butterfly labs sc 65nm devices with the addition of the 28nm
  363. device imanufacturers trumping their detection
  364. - Remove compilation warnings. Cast overly-specific #def'd values to the
  365. destination type, unsigned int.
  366. - Basic ability to compile and run on FreeBSD 10. Only tested compilation with
  367. avalon, avalon2, bflsc, and icarus; only tested functionality of bflsc and
  368. icarus.
  369. Version 4.9.0 - 16th December 2014
  370. - Minor fix
  371. - Fix MM41 voltage setting
  372. - Fix the default settings of new module
  373. - Count non matching stratum as a hw error on ava4
  374. - Fix ava4 build incompatibilites and missing write config parameters
  375. - Use strcasecmp for device matching in usbutils in case of subtle manufacturer
  376. changes
  377. - Add manufacturer and product definitions for ava4
  378. - Cosmetic ava4 change
  379. - Cosmetic ava4 message fixes
  380. - Add sanity check for NULL data being passed to usb_perform_transfer
  381. - All write errors should be treated as fatal for ava4 devices
  382. - Change initial fan start speed, mins and max for avalon4 to ensure fan starts
  383. spinning but can go lower RPM
  384. - Disable zero length packets on ava4 before trying to init
  385. - Add a cgpu device option to disable zero length packets and enable it for
  386. avalon4
  387. - Display ava4 stats consistent with other devices
  388. - Add ava4 to udev rules file
  389. - Fix build warnings on ava4
  390. - Fix ava4 build
  391. - Add Avalon4 support
  392. - Filter duplicate stratum shares from being submitted upstream
  393. - Do rudimentary detection of duplicate shares per device
  394. Version 4.8.0 - 25th November 2014
  395. - Allow forcing of building driver combinations with --enable-forcecombo
  396. - Put spaces between name and id in avalon2 and icarus
  397. - Relax detection of a failing ava2 to more than 1 minute and perform the test
  398. after polling for results
  399. - Cap maximum diff on ava2 in order to still get shares
  400. - Put space between device name and id to prevent device names with numbers in
  401. them confusing the display
  402. - USB write errors are always fatal so they should be treated as such on ava2
  403. - Issue a usb reset for ava2 that is not returning valid shares and then drop it
  404. if it persists for over a minute
  405. - Process share results without a result value
  406. - Damp out hashrate displayed for antminer USBs
  407. - Add voltage and speed where relevant to antminer USBs
  408. - Don't estimate time on any antminer usb during a timeout
  409. - Return icarus nonce ok only when the nonce size matches the device or more
  410. - Don't discard old workids until we cycle back to them on antusb and look for
  411. more nonces in the buffer
  412. - Adjust ant usb timing for queued work
  413. - Use a cyclical list for the ant queued work
  414. - Mask and limit workid for antusb and dont clear buffer
  415. - Check the nonce on the worked item, not the submitted work
  416. - Skip over unfinished work that we can't free in ant usb
  417. - Use a workid and array if possible for the small ant usb work queue
  418. - Create an array for antworks for antminer usb devices
  419. - On U3 calculate hashrate purely on shares, not timeouts
  420. - Add switches for AU3
  421. - Adjust icarus wait timeout according to device
  422. - Differentiate U3 from U1/2 as a separate driver with different parameters and
  423. adjust timing accordingly
  424. - Skip ANUs detected in rock detect
  425. - Try U3 after trying other icarus options
  426. - Add rudimentary ANU voltage setting support for U3
  427. - Fix ignoring unprefixed v6 address in api allow list
  428. - Fix minor typos in Spondoolies SP10 and SP30 drivers
  429. - Implement a basic rock_flush function to discard the base work we are rolling
  430. work from.
  431. - Task_no for rockminer from the nonce bin should simply be masked
  432. - Change rbox default correction times to 5 in a revised frequency order
  433. - Change default frequency on T1 to 330
  434. - Reinstate last received check and resend in rockminer, being more lenient at 2
  435. seconds to allow for dither errors at 1
  436. - Roll work for the rbox when possible
  437. Version 4.7.1 - 4th November 2014
  438. - Selectively yield on dropping a lock only on single CPU platforms
  439. - Make it impossible to configure in more than one device that is meant to be
  440. standalone. Add more information to configure help, along with comments for new
  441. drivers.
  442. - Add warning against system libusb in configure help
  443. - stratum_rthread sleep only 3s when all the pool have disconnected
  444. - Filter responses that don't have a result
  445. - Implement support for pool ping and json integers of zero in getversion and
  446. ping
  447. - Fix segfault when writing config with hashratio built in
  448. - Save pools in priority order at time of writing config
  449. - Set the correct flag for close on exec for sockets
  450. - Suspend stratum on removing a pool
  451. - Set CLOEXEC on sockets on linux
  452. - Drivers that take a diff should specify a max diff or it is assumed they don't
  453. support one so set max_diff to 1 if unset
  454. - Send hfa generic frame only if voltage was specified on the command line for
  455. that device
  456. - Set hashfast voltage settings only when really needed
  457. - Hashfast voltage support
  458. - Increase max diff on sp30 to 1024
  459. - Reset ipv6 flag to false in every api-allow loop
  460. - undeclared identifier 'IPV6_ADD_MEMBERSHIP' fix for apple
  461. - two back temps spondoolies2
  462. - two back temps spondoolies
  463. - correct suggest_difficulty json rpc call
  464. - Add more usb3 hub identifiers for windows
  465. - Set driver max diff to large value if unset
  466. - Wake gws on get queued
  467. - Implement blacklisting of attempting to match known products from ones without
  468. identifiers
  469. - Fix hfa driver building without libcurl
  470. - Enable building libusb without udev
  471. - Fix off by one calculation error in sp30 leading zeroes
  472. - Send correct diff work to sp30 for hashmeter to be correct
  473. - Do the sleep in spondoolies_queue_full_sp30 after dropping the lock
  474. - Minor tidy in sp30 driver
  475. - Fix sp30 warnings
  476. Version 4.7.0 - 14th October 2014
  477. - Implement generic inet_pton for windows
  478. - Fix warnings
  479. - Fix bulk of remaining style in blockerupter.c
  480. - Tidy style in blockerupter.h
  481. - Tidy bulk of style in blockerupter.c
  482. - Fix missing minimum diff setting for blockerupter
  483. - Fix unused variable warnings
  484. - remove unnecessary sleep; fix potenital div by 0 errs; use min_diff in driver
  485. definition
  486. - Fix coding style
  487. - Make the sp30 hashrate meter based on valid share generation
  488. - Change default max queue back to 1 in line with speed of most current asic
  489. controllers
  490. - Change diff limits to values suitable for sp30
  491. - Add pool number to response from addpool to the API
  492. - Make the restart and quit API commands valid json responses
  493. - Fix number of nos
  494. - Add option to set clock ('--bet-clk X' actual clock is (X+1)*10 )
  495. - compatible with X24 board
  496. - Fix error when using v6 without mask in api-allow
  497. - Support ipv6 multicast
  498. - Set min_diff to 1
  499. - Allow arbitrary clamping of lower device diffs for slow controllers by driver
  500. - Don't set default fan to max on hashratio
  501. - The 2nd read never gets anything on ava2 so remove it entirely and just return
  502. an error if we are out of sync
  503. - Implement support for mining.suggest_difficulty
  504. - Fix client ip address output
  505. - Free addrinfo garbage
  506. - Remove the brackets when using v6 pool address
  507. - Add ipv6 support for api listen
  508. - Avalon Nano: Add support Avalon Nano usb miner
  509. - fix bug in setdiff
  510. - limit minimum diff to 64
  511. - Add BlockErupter Driver
  512. - Avalon2: display currect max temperature on statline
  513. - Remove unused variable
  514. Version 4.6.1 - 20th September 2014
  515. - Throttle bflsc28 devices when they hit the overheat limit
  516. - Add whitelisting of firmware used in final bflsc28 products
  517. - API.java - remove lowercase of all data sent
  518. - Avalon2: Add 3 bytes nonce2 support
  519. - Avalon2: MM needs n2size length <= 4
  520. - Use fan min as fan speed when run with --avalon2-fixed-speed
  521. - Clear the pool submit fail bool after adding shares to the stratum hashtable
  522. to minimise window the share is not in the table
  523. - api-example unlimited socket works
  524. - Add custom strcasestr and use custom gnu type functions in bflsc
  525. - Fix windows build of bflsc driver
  526. - Fix possible deref in bflsc28
  527. Version 4.6.0 - 7th September 2014
  528. - We should not be checking for pool_unworkable in cnx_needed as it is keeping
  529. stratum connections open on unused pools
  530. - Properly handle lack of input when adding pool via menu
  531. - Allow workers without passwords
  532. - minion - increase max chip number
  533. - Avalon2: add more comments on Avalon2 options
  534. - Avalon2: add polling delay option, long coinbase support, LED status on API,
  535. change overheat from 88 to 98
  536. - minion - add a ' before non-zero core error counts
  537. - minion - hidden optional per core nonce stats
  538. - bflsc28 - clock is hex
  539. - bflsc28 - allow setting clock and volt from the API ascset command
  540. - bflsc28 - add chip count to stats
  541. - bflsc28 stats
  542. - Simplehacks to better serve bflsc28
  543. - Only use one hashtable for bflsc28 work queued
  544. - Copy back the buffer after we've stripped the inprocess field on bflsc
  545. - Parse results for BMA based on uid and remove work from the queue when found
  546. - Strip out the inprocess details from bflsc results if it exists
  547. - Create a hashtable of work by uid as it's accepted by BMA
  548. - Add some rudimentary values for BMA sleep times
  549. - Fix various errors in queueing work for bflsc28 and limit job queueing to 10
  550. to fit within a usb frame
  551. - Create preliminary work queueing for bflsc28 using jobs of up to 20 at a time
  552. by rolling work where possible
  553. - Convert all bflsc transfers to the full 512 bytes
  554. - Don't mistake bflsc28 for fpga
  555. - Do initial detection of bflsc28 devices
  556. Version 4.5.0 - 29th July 2014
  557. - Fix windows build for hashratio and ava2
  558. - Demote bad checksum message in cointerra driver but allow message to still be
  559. parsed since it won't allow existing firmwares to work otherwise
  560. - Reorder and document the configure options
  561. - Merge https://github.com/KnCMiner/cgminer into knc
  562. - Change default voltage on ava2 to 0.666V because Satan
  563. - Enable combined building of avalon2 and hashratio
  564. - Fix stratum embedded fpgas to not duplicate work with other devices
  565. - Implement smarter PID type fan control for ava2 allowing more generous
  566. temperatures and far lower fan speeds for optimal power and noise usage. Adjust
  567. default frequency to 450 as per recommendation.
  568. - Fix various warnings in ava2
  569. - Go back to polling design since async will not work for ava2 and fix various
  570. read design errors
  571. - Fix error in 2nd read functions for av2 and hro
  572. - Correct init and read sequence for ava2, and convert from a polling mechanism
  573. to a separate read thread
  574. - Initial commit of ava2 conversion to direct USB
  575. - Display frequency and voltage with ava2 on the statline
  576. - Store fan percentage and display temp and fan percent for ava2
  577. - Set avalon2 frequency and voltage to appropriate defaults if none are
  578. specified on the command line
  579. - Demote some ava2 messages that don't need to be errors and remove unused works
  580. array
  581. - Fix broken fan logic for ava2
  582. - Fix hexdump on 64bit
  583. - rockminer frequency is between 200 and 400 MHz
  584. - fix jansson include path in cgminer-api compile instructions
  585. - Remove requirement for ifndefs for avalon2 from the generic driver work
  586. function
  587. - Fix hashratio device name
  588. - Make submit_nonce2_nonce return whether the share was valid or not
  589. - Reinstate missing necessary init sequence for hashratio
  590. - Handle disconnected hashratio devices
  591. - Add hashratio frequency command line
  592. - Fix stratum updates not being passed to hashratio devices and clean up
  593. - Move to updated avalon2 type driver model for hashratio device
  594. - Initial import and conversion of hashratio driver to direct USB
  595. - Increase the internal buffer for API response, as "stats" command response
  596. can grow greater than 8K
  597. - Detach test pool thread only if we have a blocking startup
  598. Version 4.4.2 - 17th July 2014
  599. - Remove the use of the pthread_tryjoin_np which is currently unimplemented on
  600. many platforms
  601. - Fix processarg parameters loaded from a config file not being saveable
  602. - We only use the jansson in our source tree so no need for special case
  603. handling of older versions
  604. - Upgrade jansson to 2.6
  605. - Only clear sockbuf if it's been allocated
  606. - Fix missing osm-led-mode support in write config
  607. - Deal with nanosecond overflow in both directions on both addition and
  608. subtration of timespecs
  609. - Rename sp10 driver internally from spondoolies to sp10
  610. - minion - add a 2nd (optional - disabled) reset test
  611. - production stats added, reset queue added
  612. - minion - correct led ghs2 choice
  613. - minion - correct ghs2 display
  614. - minion - reset the led counter when doing a chip RSTN full reset
  615. - minion - don't reset the led counter for an SPI reset
  616. - minion - led per chip and use all time average
  617. - minion - report spi error counts and settings in stats
  618. - minion - undeclared fix
  619. - minion - chip power cycle option
  620. - minion - record 0xff error history and reduce screen output
  621. - minion - reset on result or fifo 0xff
  622. - minion - clarify the 0 value of spireset
  623. - minion - make SPI reset more configurable
  624. - minion - make the SPI reset ms sleep a parameter and API settable
  625. - sp10 sensors
  626. - sp30
  627. - minion - led+more api setting
  628. - Avoid blocking all pool testing if one pool fails to ever init
  629. - There is no point storing the hints addrinfo in struct pool
  630. - minion - 'reset' SPI when getting errors
  631. - initialise more pool values in benchmark
  632. - minion - auto adjust freq
  633. - merge upstream frequency changes
  634. - icarus - timing history in its own function
  635. - rbox - add lotsa stats, tidy up a bit more
  636. - Fix an off-by-one.
  637. - icarus - detect stat should be LOG_DEBUG
  638. - icarus - tidy up rbox code, remove statics, and add rocketbox
  639. - minion - do an early reset to clear the chip status
  640. - minion - use descriptive names for the list types
  641. - Avalon2: automatic adjust fan speed, using crc16 on job_id compare, turn on
  642. the led by API, detect twice when start, remember the last stratum message
  643. increase the hashrate, add cutoff option
  644. - fix AntS1 breakages from AntS2 changes
  645. - minion - disable dup nonce check
  646. - minion - add an ioseq number for each ioctl to simplify work ordering
  647. - minion - redo old work expiry based on txrx order
  648. - minion - more work stats, minimise queued work, free flushed queued work
  649. - minion - allow resetting a chip via the API
  650. - minion - correct 'WQue Count' in stats
  651. - minion - delay after reset, reset history also, add dups to api stats
  652. - noncedup - give access to the internal stats
  653. - minion - increase reset to 75%
  654. - minion - dup checking, disable reread by default and extra ioctl debugging
  655. - minion - always check the chip queue before queuing new work
  656. Version 4.4.1 - 21st June 2014
  657. - Move icarus driver to being seen as an asic
  658. - Clear usb reads on each pass through icarus detect to hopefully prevent false
  659. positives for detecting rboxes
  660. - Clean up pool failure and failover code for stratum
  661. Version 4.4.0 - 16th June 2014
  662. - Tidy unused rockminer variables
  663. - Tidy rockminer defines
  664. - Make rockminer driver compatible with other icarus drivers being present
  665. - Import basic rbox driver
  666. - minion - add optional (on) GPIO chip selection
  667. - Clear the pool idle flag in the pool test thread
  668. - CoreFmatch in cointerra should be a uint16
  669. - Display error message if we receive one on share rejects
  670. - Allow zero length strings to be passed to valid_hex
  671. - delete unused roundl definition
  672. Version 4.3.5 - 10th June 2014
  673. - Cointerra driver updates.
  674. - Sleep before retrying in the test pool thread after a pool has died
  675. - Use valid_ascii testing for job_id since it need not be hex only
  676. - Only show slow/down message till pool is flagged idle
  677. - Do some random sanity checking for stratum message parsing
  678. - Keep looking for when a pool comes to life at startup and touch the logwin so
  679. the message is not invisible
  680. - Fix no libcurl build
  681. - Added Drillbit Thumb to udev rules.
  682. - Avoid dereference on getting API stats on partially initialised HFA instances
  683. - A1: add support for updated product variants, small fixes
  684. - Add one more usbutils fix
  685. - Convert uses of usbutils memcpy to cg_memcpy
  686. - Add a sanity checking memcpy function which checks for overflows
  687. - minion - count force use reread
  688. - minion - add a disabled ioctl() test
  689. - minion - add more checking of SPI results for corruption
  690. - minion - optional (disabled) ioctl() debug
  691. - Increase S1 overheat to 75 degrees C
  692. - Add ruby api-example to API-README
  693. - minion - allow core selection at runtime
  694. - API - lcd all-in-one brief summary
  695. Version 4.3.4 - 25th May 2014
  696. - Add support for 2 nonces per block in spond driver
  697. - Increase timeout on reset in cta driver to 5 seconds
  698. - Increase max diff on spondoolies driver slightly to be well below spi comms
  699. limitations
  700. - Use the active contents lock and safe list iteration within the linux usbfs
  701. code
  702. - Add Ruby Api Example
  703. - Automatic detect the small miners
  704. - Update default modules from 3 to 4
  705. - Fix the temp max. we should use currect max temp
  706. - add avalon2-cutoff options
  707. - Enable the cutofftemp to Avalon2. ignore longer coinbase and longer merkles
  708. stratum
  709. - Fix the diff value used on MM firmware
  710. - Mark pool as idle if stratum restart is failed
  711. - Add hacky workaround for double list removal race in libusb
  712. - Make the work given in benchmark mode deterministic on a per-device basis
  713. - Rework the benchmarking code to use a deterministic set of work items with a
  714. known number of diff share nonces at regular spaced intervals
  715. - minion - restrict nonce read result size to ioctl() limit
  716. - minion - must check temp when overheated
  717. - minion - idle chips that hit >100C until back to 80C
  718. - minion - report the chip/reg when aborting due to an invalid ioctl() size
  719. - minion - all freq in Mhz but only convert when used
  720. - minion - remove unused ioctl debug
  721. - minion - command queue is now larger
  722. - minion - check rolled in stale work cleanup
  723. - Work stats should be based on device_diff not work_difficulty since non-shares
  724. haven't been filtered out yet
  725. - Prevent a segfault when writing a config file containing 'rotate' option
  726. - minion - comment out HW debug message
  727. - minion - roll work to reduce CPU
  728. - minion - report init_freq in stats
  729. - api - howoldsec is only used for USB
  730. - minion - allow setting the frequency
  731. - minion - disable iostats by default since it slows down mining
  732. - minion - define frequency value table
  733. - minion - report temp/cores/freq and handle temp formatting
  734. - minion - item is undefined
  735. - Rationalise diffs stored in the work struct and document them to avoid further
  736. confusion
  737. - Add basic API stats for nfu drivers to see how many submits each chip returns
  738. - Add output direction for the EN0 pin on nfu driver
  739. - Support power management optimisations in newer nf* firmware
  740. - Support variable numbers of chips with NFU and BXM drivers
  741. - Identify number of chips in nanofury devices and change name accordingly
  742. - Rename nf1 driver to nfu in anticipation of support for more chips
  743. - Make hashfast reset counter rise on old instances when inheriting the value on
  744. new ones
  745. Version 4.3.3 - 3rd May 2014
  746. - Fix typo
  747. - Work should be freed when aged, fixing a massive memory leak for bxf devices
  748. - miner.php fix single rig summary/config field formatting
  749. - miner.php fix single rig total formatting
  750. Version 4.3.2 - 2nd May 2014
  751. - Fix accounting bug with nrolltime drivers
  752. Version 4.3.1 - 2nd May 2014
  753. - upgrade some int to int64_t to avoid overflows in reporting
  754. - Make reconnection messages more explanatory
  755. - Stratum client.reconnect require matching URL
  756. - Fix memory leak in submit_noffset_nonce
  757. - Clean up any work that may not have been used in the work scheduler
  758. - Avoid unnecessary deref now that it's done within discard_work
  759. - Clean work pointers after one way usage functions
  760. - Avoid unnecessary total_work_inc in generating local work
  761. - Cosmetic fixes
  762. - Fix idle bug, when redirected client can't auth
  763. - Rename spond temp rate to asics total rate
  764. - Build fixes
  765. - Set the unique id only for usb devices with serial strings longer than 4 chars
  766. long
  767. - Use usb serial strings as unique id if devices have them
  768. - Discretely identify the onestring miners as OSM
  769. - Add bxf debugging option and osm led modes
  770. - A1: modularize board selector / add initial CCR support
  771. - A1: cleanup tca9535 logging
  772. - A1: fix and extend PLL parameters
  773. - A1: clean up compile warnings
  774. - A1: use real level in hexdump
  775. - Add identification for onestring miner variants
  776. - Avalon2: Parser the power good signal
  777. - driver-avalon2: this functions used on detect, which don't have thr setup yet
  778. Version 4.3.0 - 18th April 2014
  779. - Put sleep in spond hash instead of queue full function
  780. - Remove unused function for when compiled without curses
  781. - Fix typo
  782. - Add temperature rate, front, rear and device temperature to spond API output
  783. - Limit bxf sleep in bxf_scan to 100ms minimum for strings of many chips
  784. - -Werror=format-security error on driver-bitmain.c
  785. - Fix parameters passed with getblockhash
  786. - Check the block hash with the proper command when looking for orphan chains
  787. - syslog requires a facility ... in more than one place
  788. - Shuffle windows headers included
  789. - Adjust the bxf sleep time according to the number of chips detected
  790. - Fix off by one error in bxf chip count when adjusting device size
  791. - Recalloc correct pointer
  792. - Make instructions associated with winusb error even more explicit
  793. - Add midsing headers to cgminer source in Makefile
  794. - Trivial style changes to mg proto parser
  795. - Trivial style and warning clean ups on spondoolies driver
  796. - Merge spondoolies driver patch
  797. - Call any BXF device with 3-6 chips reported HXF
  798. - Avoid derefrence when calling statline before on hfa device during init
  799. sequence
  800. - Calloc the info structures even on failed hfa reset to prevent later possible
  801. dereference
  802. - Load all hfa devices based on identification alone and defer init sequence
  803. till mining thread init sequence to allow all devices to be recognised rapidly
  804. but each device initialisation not delay others
  805. - Do not do thread shutdown unless thread init succeeded
  806. - Remove unnecessary check for thread_prepare function
  807. - Recognise variations on BXF based on chip value returned in responses
  808. - Provide helper function for recallocing memory
  809. - syslog requires a facility
  810. Version 4.2.3 - 3rd April 2014
  811. - Decay the per device hashrates when only the watchdog is calling the hashmeter
  812. - Fix parsing of config files failing on custom parsing
  813. - Allow an arbitrary number of chips in the BXF driver, showing results from
  814. each chip in the API and identify the hexfury, naming it HXF
  815. - Disable toggling display by default and offer a --widescreen option to have
  816. all the information on an extra wide display.
  817. - Use OPT_WITH_CBARG for all custom parsing functions to allow their values to
  818. be written generically when writing the config file from the menu.
  819. - Provide a ccan variant OPT_WITH_CBARG that assigns the arguments passed as a
  820. string and then performs the callback function on the string.
  821. - Define strings to store special option parsing parameters leaving no
  822. OPT_WITH_ARG missing args
  823. - Correct the writing of special case options to the config file
  824. - Provide support for writing anu freq from menu write option
  825. - Update to diver-avalon2.c
  826. - Generalise a lot more of the command line options simplifying the write config
  827. function and making it write far more values unaided
  828. - Use the general opt_set_charp functions for setting api parameters
  829. - Json escape any strings written to the config file
  830. - Store standard charp options when writing config files
  831. - Add support for all the integer range options when writing the config file
  832. from the menu
  833. - Remove the --device and --remove-disabled options which don't work in a
  834. meaningful way any more
  835. - Make the bxf bits configurable on the command line
  836. - Provide a --btc-sig option to optionally add a custom signature to the solo
  837. mining coinbsae
  838. - Compact gbt solo extra data and store the length, allowing it to be variable,
  839. leaving room for a signature
  840. - miner.php - Kano summary Pool Acc/Rej should be only work submitted
  841. - miner.php add best share and gen formatting for pool summary
  842. - miner.php - remove BGEN/GEN eval() errors from the web log
  843. - miner.php allow optional fields when gen is disabled
  844. - miner.php dont format missing gen fields
  845. - miner.php make Summary a custompage
  846. - miner.php allow uers and system lists of customsummarypages and add more
  847. examples
  848. - Fix getwork share submission
  849. - Cosmetic fix to udev rules
  850. - Put WU on the hashrate status to compact lines further
  851. - miner.php show api/rig errors at the top of a customsummarypage
  852. Version 4.2.2 - 29th March 2014
  853. - Minor correctness fix for unnecessary free
  854. - Clean up various curl build issues
  855. - allow url based config files
  856. - Frequency only needs 3 digits for cointerra statline
  857. - Use the serial number as unique_id for cta display
  858. - Make it possible to enable/disable the status window from switching via the
  859. display menu
  860. - We should not update the tv hashmeter time unless we're updating the hashrates
  861. - Add cointerra devices to udev rules.
  862. - Use hashfast unique id instead of number since the unique id is displayed
  863. - Remove displayed space
  864. - Left align the displayed unique id
  865. - Use the hashfast opname as its unique identifier
  866. - Display BF1 serial number as its unique identifier
  867. - Display a unique identifier instead of a number if the device has one
  868. - Use an alternating status display to return to a compact width of 80
  869. characters, allowing more information to be displayed.
  870. - No need for looking for khash hashrates in summary any more
  871. - Fix two potential minor mem leaks
  872. - Fix memory leaks in setup and generate work for gbt solo.
  873. - Fix off by one malloc size error
  874. - Fix memory leak in update_gbt_solo
  875. - Put sanity check on decay_time to prevent updates with no time
  876. - Add 3 rolling average hashrates to API output for summary and devs.
  877. - Use the extra status screen real estate better, displaying rolling 1/5/15min
  878. average hashrates as well.
  879. - Revamp the ageing crufty hashmeter code to have proper exponential decaying
  880. values and store rolling 1/5/15min hashrates.
  881. - Increment total_work under control lock.
  882. - Trivial variable reuse
  883. - Add support for other usb3 hubs on windows
  884. Version 4.2.1 - 24th March 2014
  885. - Fix various ava2 build issues generically
  886. - Minimise the amount of heap memory allocations/frees when submitting gbt
  887. shares.
  888. - Make varint in gbt submission a stack object.
  889. - Fix big endian problems with gbt submissions.
  890. - Fix 32bit overflow on relative diff shown.
  891. - ants1 - stop results read hard looping
  892. - ants1 - slow down mining if overheat occurs
  893. - miner.php allow gen before (bgen) and after (gen) grouping
  894. - Change default solo mining to failing when no btc address is specified.
  895. - Use upgrade cglock variants in get_gbt_curl
  896. - Provide a cg_uilock to unlock the intermediate variant of cglocks.
  897. - Use the one curl instance for all gbt solo operations, protecting its use with
  898. a bool set under gbt lock.
  899. - Only start block detection with gbt solo if setup succeeded
  900. - One less block detection message
  901. - Toss out the curl handle after each solo poll
  902. - Don't reuse any curl handles for solo mining and break out of the lp thread if
  903. the pool is removed.
  904. - Make sure to only start the lognpoll thread once on gbt solo.
  905. - Don't keep RPC connections open for solo mining since bitcoind doesn't like
  906. having many persistent connections.
  907. - GBT solo pools should be considered localgen pools.
  908. - miner.php - speed up formatting and allow calc on gen fields
  909. - Always show the address we're solo mining to to avoid confusion for when no
  910. address is set.
  911. Version 4.2.0 - 18th March 2014
  912. - Fix missing htobe16 on windows and meaningless >u32 string warning.
  913. - Software ntime roll for all hashfast devices.
  914. - Silence harmless warning.
  915. - Drop a failed restart icarus device to allow it to be rehotplugged if
  916. possible.
  917. - Work with more than one transaction.
  918. - Kill gbt solo pools that don't respond to the gbt request 5 times
  919. sequentially.
  920. - Fix ser_number for no remaining val byte.
  921. - Create a work item and stage it when updating the gbt solo template to allow
  922. new block detection and restart code to work.
  923. - Test block hash as well as block height when solo mining to ensure we haven't
  924. been mining on an orphan branch.
  925. - Fix transaction processing for gbt solo.
  926. - Encode height using integer varint format.
  927. - Make new block detection message not show in gbt solo from test_work_current
  928. - Add block detection via getblockcount polling in gbt solo and update gbt
  929. template every 60 seconds.
  930. - Iterate over transactions twice to malloc only once when copying all the
  931. transaction data.
  932. - Update solo coinbase regularly and submit as gbt work
  933. - Only show merkle hashes for solo mining in debug mode.
  934. - Set correct flag for solo work.
  935. - Generate gbt solo work emulating stratum work construction.
  936. - Set the diff as a double sdiff from gbt solo data.
  937. - Move swork.diff out of the stratum work section to be shared as sdiff.
  938. - Generate a header bin from gbt solo as per the cached stratum one.
  939. - Store strings similar to stratum's when decoding gbt solo
  940. - Avoid allocing and freeing stratum strings that should be fixed length.
  941. - Run parser through detect_stratum after stratum+tcp:// is force added
  942. - Remove unnecessary header length calculation for stratum header binary and
  943. only binary convert the correct length of the header.
  944. - Share more fields between stratum and gbt
  945. - Share coinbase_len variable b/w stratum and gbt and setup more gbt solo
  946. parameters.
  947. - Generate a valid coinbase and set nonce2offset for gbt solo
  948. - Move scriptsig header bin conversion to setup gbt solo
  949. - Create our own custom scriptsig base.
  950. - Add helper functions for creating script signature templates and beging
  951. building template.
  952. - Do gbt solo decoding under gbt lock.
  953. - Add more gbt variable decoding from gbt solo information.
  954. - Store all the transaction data in binary form when using GBT
  955. - When setting up solo mining, check validity of bitcoin address against
  956. bitcoind
  957. - Make pooled GBT mining use merkle bin optimisations slated for solo mining.
  958. - Abstract out the merkle bin calculation for gbt solo
  959. - Implement efficient merkle tree base from solo GBT information.
  960. - miner.php custom formatting and row counter '#'
  961. - Drillbit: Fix for underestimating hash rate from Bitfury devices
  962. - Send per-core hashrates at regular ~5min intervals back to cta devices.
  963. - Calculate the cta per core hashrate at 5 minute intervals.
  964. - Check the bits of the correct core in cta bit count.
  965. - Display the bit count along with the bitmap for each cta core in the API stats
  966. output.
  967. - Store and display the per core hashrate on cta relative to each work restart.
  968. - Decrease the time we wait for unsetting a core on the cta bitmap to correspond
  969. with the lower max diff of 32.
  970. - Set max diff on cointerra devices to 32 which is still only 11 shares per
  971. second but allows for earlier confirmation of per core hashrates.
  972. - Keep track of when the last restart and work updates were triggered and
  973. provide helper functions for knowing the time since then.
  974. - hashfast make api stats field names unique
  975. - Fix gcc longjmp warning in api.c
  976. - Add a per-core hashrate to the cta API stats.
  977. - miner.php support edevs and estats
  978. - API - put edevstatus where it was supposed to be
  979. - Icarus - allow timing mode to work with ANU and not slow it down
  980. - drillbit - remove warnings
  981. - drillbit - minor code tidy up
  982. - Drillbit: Change language around 'void' to warning about limiter disabled
  983. - Drillbit: Fix accidental over-counting of HW errors
  984. - Drillbit: --drillbit-auto parameter for tweakable custom tuning of ASIC speeds
  985. - Drillbit: Output warning if board reports void warranty
  986. - Drillbit: Add Avalon & drillbit-autotune notes to ASIC-README
  987. - Drillbit: Limit work sent out to 8 units in a single pass, was DoSing a full
  988. double scroll
  989. - Drillbit: Move drillbit_empty_buffer calls to only when errors occur, were
  990. limiting performance on Windows
  991. - Fix Windows bug with libusb_reset_device returning SUCCESS for disconnected
  992. device
  993. - Drillbit: Fix some warnings
  994. - Drillbit: Add --drillbit-autotune option for device to dynamically alter clock
  995. speed
  996. - Drillbit: Fix typo in previous commit
  997. - Drillbit: Remove default config in cgminer, rely on defaults in firmware
  998. - Drillbit: Combine split USB transfer for sending new work, reduce overhead
  999. - Drillbit: Add support for protocol V4, with device-agnostic board
  1000. configuration data
  1001. - Drillbit driver: Add support for Avalon-based Drillbit miners
  1002. - API - add edevs and estats - to only show enabled devices
  1003. - Check device data exists on a hfa instance before trying to reinit it.
  1004. - Print off what quadrant regulator failed if known in hfa driver.
  1005. - Reset all the stats on autovoltage complete in cta driver.
  1006. - Use correct diff instead of diffbits in cta driver.
  1007. - Whitelist all firmwares <= 0.5 on hfa for software rolling of ntime.
  1008. - Avoid a memory leak by reusing the ntime field when rolling stratum work.
  1009. - Clear the pipe bitmap on cta only when no share has occurred for 2 hours
  1010. instead of 1.
  1011. - Cta share_hashes should be added, and we can base it on device wdiff instead
  1012. of pool work difficulty for more accurate hashrates.
  1013. - Since the device runtime is now reset, the Raw hashrate entry in the cta API
  1014. output is no longer meaningful.
  1015. - Look for autovoltage returning to zero on cta driver and reset stats at that
  1016. point since the hashrate is unreliable till then.
  1017. - ants1 - cgminerise applog calls
  1018. - Default to stratum+tcp:// on any urls that don't have a prefix instead of
  1019. http.
  1020. - Trivial cta style changes.
  1021. - ants1 - fix/enable temperature checking and remove unneeded temp_old
  1022. - ants1 - move local cgpu variables to info structure
  1023. - ants1 use a klist to store work and copied work
  1024. - Simplify dramatically the cross-process cgminer locking through use of flock
  1025. instead of sysv semaphores.
  1026. Version 4.1.0 - 8th March 2014
  1027. - Correct fix for dev start time being adjusted for stat zeroing.
  1028. - Make per device stats work for average after a stat zeroing.
  1029. - Add an hfa-options command line that allows the clockspeed to be chosen per
  1030. device by name comma separated, with a function that can be expanded with more
  1031. options in the future.
  1032. - Off by one drv_rolllimit check against jobs
  1033. - Free the work that may be lost, leaking memory, in a failed hfa_send_frame
  1034. - Roll the ntime for work within the hfa driver for firmware we know doesn't do
  1035. it internally as an optimisation.
  1036. - Export the roll_work function to be usable by driver code and make it
  1037. compatible with rolling stratum work.
  1038. - Make opt_queue be respected as a maximum value for staged items.
  1039. - Disable mistakenly enabled lock tracking.
  1040. - api version update for HEX32
  1041. - api.c - HEX32 type needs quotes
  1042. - Disable the MAX_CLOCK_DIFF check for newer hashfast firmwares since it's not
  1043. required.
  1044. - Store the hardware and firmware revision in the info struct for easy use in
  1045. the hfa driver.
  1046. - Only decrease the hfa clock rate if the device has been running for less than
  1047. an hour before dying.
  1048. - Change lack of op name response message in hfa driver
  1049. - Check for lost devices at every write/read in hfa_detect_common
  1050. - Make bxm bits configurable.
  1051. - Move avalon2 options to ~alphabetic position in help.
  1052. - Do a shutdown routine on bxm close.
  1053. - Provide support for 2 chips in libbitfury sendhashdata and enable the 2nd chip
  1054. on BXM devices.
  1055. - Remove unnecessary opayload and newbuf members of bitfury info struct.
  1056. - Add an spi add fasync command.
  1057. - Cope with older hfa firmware not even responding to op_name.
  1058. - Forcibly kill everything silently with an exit code of 1 should we fail to
  1059. cleanly shut down and use a completion timeout for the __kill_work in
  1060. app_restart.
  1061. - Make __kill_work itself also be a completion timeout.
  1062. - Generalise more of libbitfury for more reuse in both nf1 and bxm drivers.
  1063. - Remove redundant init components of bxm driver.
  1064. - Set default osc6 bits on bxm to 50
  1065. - Enable the transaction translator emulator for bxm devices and use a dummy spi
  1066. tx the size of a normal payload.
  1067. - Store usb11 and tt flags as booleans in cgusbdev allowing them to be
  1068. discretely enabled as well as detected by the device data.
  1069. - Add bxm scan function and check spi txrx returns only as much as sent.
  1070. - Add init sequence to bxm detect one.
  1071. - Add a bxm specific txrx function for spi transfers.
  1072. - Add bxm close to bitfury shutdown switch.
  1073. - Add reset/purge/cshigh/low sequence to bxm init
  1074. - Add bitmode init to bxm open sequence.
  1075. - Add initial bxm opening sequence for detect one.
  1076. - Add identifiers for bxm bitfury devices.
  1077. - Clean up parse_method
  1078. - More gracefully break out of parse_notify on a corrupted hex string error,
  1079. checking the return value of all hex2bin conversions and being consistent with
  1080. using stack memory. Fix an unlocking error in cases of failure.
  1081. - AntS1 - add detection information to usbutils
  1082. - Enable Bitmain Ant S1 code and make it conform to cgminer requirements
  1083. - Make the cointerra displayed hashrate based on valid share generation.
  1084. - Convert and update values shown in the cointerra api output.
  1085. - Export the api_add_int16 function.
  1086. - Use a custom mystrstr function in cointerra driver.
  1087. - Add api_add_int16 to API functions.
  1088. - Add support for Bitmain Multi Chain and Single Chain and optimize the
  1089. efficiency
  1090. - Add support for bitmain devices
  1091. - Perfect function of BitMain Multi Chain
  1092. - Add support for Bitmain Multi Chain and Single Chain and optimize the
  1093. efficiency
  1094. - Add support for bitmain devices
  1095. Version 4.0.1 - 28th February 2014
  1096. - Refresh the log window on pool failure message at startup.
  1097. - Rework the pool fail to connect at startup to not get stuck indefinitely
  1098. repeatedly probing pools with new threads and to exit immediately when any key
  1099. is pressed.
  1100. - Use an early_quit function for shutting down when we have not successfully
  1101. initialised that does not try to clean up.
  1102. - Add more information to a hfa bad sequence tail event.
  1103. - Increase the work queue at the top end if we've hit the bottom as well.
  1104. - Set the work generation thread high priority, not the miner threads.
  1105. - Bringing each hfa device online takes a lot of work generation so only ever do
  1106. one at a time.
  1107. - Increase the opt_queue if we can hit the maximum amount asked for but are
  1108. still bottoming out.
  1109. - Keep the old hfa device data intact with a clean thread shutdown to allow it
  1110. to be re-hotplugged with the old information.
  1111. - Cope with the API calling hfa on partially initialised devices having no info.
  1112. - Show only as many digits as are required to display the number of devices.
  1113. - Cold plug only one hashfast device to get started, and then hotplug many to
  1114. minimise startup delays and possible communication delays causing failed first
  1115. starts.
  1116. - Send a shutdown and do a usb_nodev if hfa_reset fails.
  1117. - Null a device driver should thread prepare fail.
  1118. - Add a function for making all driver functions noops.
  1119. - Don't try to reinit a device that's disabled.
  1120. - Disable a device that fails to prepare.
  1121. - Check for lack of thread in watchdog thread for a failed startup.
  1122. - Make all device_data dereferences in the hfa driver safe by not accessing it
  1123. in statline before when it's non-existent.
  1124. - Add an option to disable dynamic core shedding on hashfast devices.
  1125. - Do not remove the info struct on a failure to hfa prepare.
  1126. - Detect an hfa device purely on the basis of getting a valid header response to
  1127. an OP_NAME query, leaving init to hfa_prepare which will allow multiple devices
  1128. to start without holding each other up at startup.
  1129. - Store the presence and validity of opname in the hfa info.
  1130. - api - buffer size off by 1 for joined commands
  1131. - minion - clean up statline
  1132. - Only break out of usb_detect_one when a new device is found.
  1133. - Use usb_detect_one in the hfa driver.
  1134. - Provide a usb_detect_one wrapper which only plugs one device at a time,
  1135. breaking out otherwise.
  1136. - Issue a usb_nodev on a bad work sequence tail in hfa
  1137. - Read in hfa stream until we get a HF_PREAMBLE
  1138. - Add shed count to hfa API stats output.
  1139. - Display the base clockrate for hfa devices with a different name to per die
  1140. clockrates to be able to easily distinguish them.
  1141. - Use op_name if possible first with hfa devices to detect old instances and be
  1142. able to choose the starting clockspeed before sending an init sequence,
  1143. reverting to setting op name and serial number as fallbacks.
  1144. - Make hfa resets properly inherit across a shutdown.
  1145. - Don't break out of hfa_old_device early if there's no serial number.
  1146. - Fix harmless warning.
  1147. - Allow the drop in MHz per hfa failure to be specified on the command line.
  1148. - Icarus - ignore HW errors in hash rate ... and fix detection of them
  1149. - Enable the hfa shed supported feature by default.
  1150. - Add to udev rules hfa devices for firmware writing.
  1151. - Remove ENV from hashfast udev rules.
  1152. - Add a --hfa-name command that allows one to specify the unique opname for a
  1153. hashfast device.
  1154. - Ava2 decode the voltage, get the temp_max
  1155. - Set the clock rate with a work restart instead of an init when changing to old
  1156. clocks for hfa
  1157. - Set opname on hfa devices without a serial number to a hex value based on time
  1158. to not overflow the field.
  1159. - Add op name to hfa API stats output if it exists.
  1160. - Set the actual op_name in hfa devices if cgminer is choosing it itself due to
  1161. it being invalid.
  1162. - Re-init an hfa device to its old data before setting up info structures as
  1163. their sizes may change.
  1164. - Remove the usb device whenever we do a running shutdown on hfa and do a
  1165. shutdown as the imitated reinit to allow it to hotplug again.
  1166. - Reset opt hfa dfu boot after it's used.
  1167. - Comment out windows only transfer on hfa startup.
  1168. - Clean up structures unused in case of all failures in hfa detect common
  1169. - Clear all structures should we fail to hfa reset on adjusting clock on a
  1170. hotplug.
  1171. - Set master and copy cgpu hash clock rate for hfa when dropping it on a
  1172. restart.
  1173. - Set the master hfa clock speed to lower when shutting down a copy.
  1174. - Do a clear readbuf on any hfa reset in case the device has not yet cleanly
  1175. shut down.
  1176. - Increase hfa fanspeed slightly more when it's rising in the optimal range than
  1177. falling.
  1178. - Always decrease hfa clock speed on a running shutdown and don't try sending an
  1179. init frame since it will be dropped regardless.
  1180. - Match hfa devices to old ones based on OP_NAME values before serial numbers if
  1181. possible.
  1182. - Read off the OP_NAME if it exists and is supported on hfa devices, setting it
  1183. to the device serial number or a timestamp if it is invalid.
  1184. - Updated hf protocol
  1185. - Check for an amount along with no error in hfa clear readbuf
  1186. - Hfa clear readbuf can return a nonsense amount when there's a return error so
  1187. ignore the amount.
  1188. - Running resets always cause a shutdown on hfa meaning the device will
  1189. disappear with modern firmware so always kill off the threads to allow
  1190. re-hotplugging.
  1191. - Reset the hfa hash clock rate to the old one if we find an old instance, only
  1192. setting the device id in hfa_prepare
  1193. - Keep the device_id on the original zombie thread for HFA in case of further
  1194. resets.
  1195. - Break out of hfa inherit if there is no device data.
  1196. - Inherit the hfa zombie instance after the device id has been allocated.
  1197. - The list_for_each_cgpu macro will dereference when there are no mining threads
  1198. yet.
  1199. - Make hfa hotplug inherit some parameters from a previous instance if the
  1200. serial number exists and is matching, avoiding dropping the clock on all
  1201. devices.
  1202. - Per device last getwork won't work if the device stops asking for work.
  1203. - Use the share_work_tdiff function in the driver watchdogs.
  1204. - Provide a helper function for determining time between valid share and getwork
  1205. per device.
  1206. - Store last_getwork time on a per-device basis.
  1207. - Limit the decrease of hfa clock rate on reset to the default clockrate.
  1208. - Base the hfa failure time on the current expected hashrate instead of a static
  1209. 15 seconds.
  1210. - We shouldn't be trying to read from the hfa_send_shutdown function itself.
  1211. - Reset the icarus failing flag only when a valid nonce is found.
  1212. - Transferred value is corrupt on a NODEV error in usbutils.
  1213. - Set each miner thread last valid work just before starting its hash loop in
  1214. case there are delays at startup.
  1215. - Only memcopy *transferred data in usbutils if we have received only success or
  1216. a non-fatal error.
  1217. - Increase to 25 nonce ranges on icarus fail detect.
  1218. - Set icarus device fail time to be dependent on device speed to avoid falsely
  1219. detecting failure on slower AMU devices.
  1220. - Updated hf protocol header.
  1221. - Updated BE hf protocol header.
  1222. - Take into account shed cores on hfa devices when determining how many jobs to
  1223. send.
  1224. - Fix compilation error with two avalon types.
  1225. - Fix missing A1 files from distribution.
  1226. Version 4.0.0 - 21st February 2014
  1227. - Check for error from setfloatval
  1228. - Halfdelay cannot be larger than 255.
  1229. - Allow any arbitrary frequency to be specified for ANU devices and try to find
  1230. the nearest frequency when initialising it, reporting if the frequency is not
  1231. exactly as requested.
  1232. - Only show system libusb warning when appropriate during configure.
  1233. - Merge branch 'avalon2' of https://github.com/xiangfu/cgminer into
  1234. xiangfu-avalon2
  1235. - Hfa cooling remains satisfactory down to a minimum fanspeed of 5%
  1236. - Give a nodev error if we have already set nodev in hfa clear readbuf to avoid
  1237. further usb comms attempts.
  1238. - Fix missing include
  1239. - Move bitmine options to alphabetic positioning.
  1240. - bab - missed a few 'DEAD's in last commit
  1241. - bab - use 'bad' instead of 'dead' as per the screen B:
  1242. - bab - roll work if possible to reduce CPU
  1243. - Update the per die hash clock data on a running reset on hfa devices.
  1244. - Set the per die clock on hfa to the known starting base clock instead of our
  1245. requested clock rate.
  1246. - Hfa device failure can be detected within 15 seconds so we should try
  1247. restarting it sooner to avoid tripping the device's own watchdog.
  1248. - Check return result of hfa clear readbuf to minimise error messages on device
  1249. failure.
  1250. - Put MHz into cta statline description.
  1251. - Send a work restart with every shutdown message to hfa devices to clear any
  1252. work that might be stale on the next restart.
  1253. - Store the hfa hash_clock rate and display it in the statline.
  1254. - Store the maximum board temperature for hfa devices and take that into
  1255. consideration when calculating the highest temperature as well as the dies.
  1256. - A1: CoinCraft-Desk driver variant
  1257. - Initial import of Bitmine.ch A1 SPI driver
  1258. - klondike ensure stats type matches
  1259. - avalon, bab, drillbit, klondike use more screen space rather than truncating
  1260. info
  1261. - Add hashfast fanspeed% to statline display.
  1262. - Move driver statline padding to cgminer.c, expanding width of maximum
  1263. displayable statistics and window width to add more info.
  1264. - Prune old stratum shares that we've seen no response for over 2 minutes to
  1265. avoid memory leaks for pools that don't respond about some shares.
  1266. - Add warning if system libusb is being added.
  1267. - Only run ./configure with autogen.sh if extra parameters are passed to it.
  1268. - Updated cointerra features.
  1269. - Add le16toh defines for platforms that may be missing it.
  1270. - Remove modminer bitstreams from distribution and replace with a README saying
  1271. what file needs to be added if modminer build is still desired.
  1272. - Use the simplelog function from usb_list()
  1273. - Add a simplelog function that does not log date and time.
  1274. - Use a unique usb_list function displaying only pertinent information when
  1275. listing usb devices from the menu.
  1276. - Abstract out the _in_use function to take different linked lists.
  1277. - Break out of linked list loop in remove_in_use in case we've gone over the
  1278. whole list.
  1279. - Check for hfa invalid hash clockrate after other error messages.
  1280. - Detect non-responsive icarus devices and attempt a usb reset before killing
  1281. them after 2 minutes of no hashes.
  1282. - Detect non-responsive bitfury devices and try a usb reset on them before
  1283. killing their instances off after 2 minutes of no activity.
  1284. - Allow hotplug interval to be changed from the USB menu.
  1285. - Prevent recursive loop in __is_in_use linked list walking.
  1286. - Add the ability to whitelist previously blacklisted usb devices from the menu.
  1287. - Use a bool in struct cgpu to know when a usb device has been blacklisted,
  1288. avoiding blacklisting it more than once.
  1289. - bab - ensure disabled chips are counted in the screen dead chip counter
  1290. - bab - only disable the chip once ...
  1291. - bab - short work list skip disabled chips
  1292. - api.c avoid incorrect gcc warning
  1293. - cgminer -h crash fix
  1294. - Add blacklisting as an option to the USB menu.
  1295. - Add a mechanism to blacklist a usb device from its cgpu.
  1296. - Add an option to the USB menu to list all known devices.
  1297. - Add an option to send a USB reset via the USB menu.
  1298. - Add a usb_reset by cgpu function to usbutils.
  1299. - Add warning for attempting to unplug a usb device that is already removed.
  1300. - Add USB Unplug option to USB management device management menu.
  1301. - Add enable and disable USB device functions to the menu.
  1302. - Add a [U]SB menu item, initially with just statistics per device, adding
  1303. device number to the device status window display.
  1304. - Reuse the cgpu temp entry for avalon and bitfury devices, changing avalon to a
  1305. damped value.
  1306. - Store the cointerra maximum temperature in the cgpu struct as an exponentially
  1307. changing value based on the maximum temperature.
  1308. - Reuse the cgpu->temp entry for max temperature in hfa driver.
  1309. - bab - disable chips that return only bad results
  1310. - Add driver for cointerra devices.
  1311. - Add Avalon2 (2U size machine) support
  1312. - miner.php - define a default rigport (that can be changed) and don't require a
  1313. port number in the rigs array
  1314. - miner.php allow links for rig buttons in tables and allow using the 4th IP
  1315. octet if no rig name - default disabled for both
  1316. - format fix and bad variable usage fix for --benchfile
  1317. - Allow running cgminer in benchmark mode with a work file --benchfile
  1318. - ANU frequency is in MHz, not hex.
  1319. - Remove bitfury devices from the usb list on shutdown in case they have stopped
  1320. responding but have not had a fatal usb error.
  1321. Version 3.12.3 - 8th February 2014
  1322. - Put the hashfast temperature into the cgpu structure so that it shows up in
  1323. the devs API call.
  1324. - We shouldn't block on no work situations directly from the getwork scheduler
  1325. itself.
  1326. - Revert "Make the pthread cond wait in the getwork scheduler a timed wait in
  1327. case we miss a wakeup."
  1328. Version 3.12.2 - 8th February 2014
  1329. - Adjust antminer U1 timing according to command line frequency set, fixing the
  1330. need for icarus timing on the command line.
  1331. - Read pipe errors that don't clear are worth attempting to reset the usb.
  1332. - Revert "Do away with usb resets entirely since we retry on both pipe and io
  1333. errors now and they're of dubious value."
  1334. - Make the pthread cond wait in the getwork scheduler a timed wait in case we
  1335. miss a wakeup.
  1336. Version 3.12.1 - 7th February 2014
  1337. - Document new features for antminer U1 and hfa devices.
  1338. - Add support for ANU overclocking.
  1339. - Increase hfa fanspeed by more if we're rising in temp above the target than if
  1340. the temp is staying the same.
  1341. - Add debug output when get_work() is blocked for an extended period and add
  1342. grace time to the device's last valid work to prevent false positives for device
  1343. failure.
  1344. - Issue a shutdown prior to a reset command for hfa devices and lock access to
  1345. reads awaiting the response if the device is already running.
  1346. - Do not register as successful a hfa init sequence that reports the clockrate
  1347. as zero.
  1348. - Show device info in noffset nonce share above target message.
  1349. - Widen lines in top menu to fit extra large share values.
  1350. - Only show one decimal place if pool diff is not an integer.
  1351. - Show serial number as a hex value in hfa verbose startup.
  1352. - Slowly remove work even if it's not being used to keep the getwork counter
  1353. incrementing even if work is not used and as a test that pools are still
  1354. working.
  1355. - Increase the maximum diff between hfa dies to 100Mhz.
  1356. - Show which hfa die is bringing down all the others when decreasing all the
  1357. clock speeds.
  1358. - Increase the decrease when temp has increased more and we want to decrease it
  1359. on hfa.
  1360. - Give device info with share above target message.
  1361. - Allow throttling of hfa dies more frequently and increasing of speeds less
  1362. frequently.
  1363. - Wait after sending a hfa shutdown to allow the device to properly shut down
  1364. before possibly sending it more commands.
  1365. - Minimise the die clock differences in hfa to no more than 50Mhz.
  1366. - Check for when errno is set on windows as well as the windows variant for
  1367. errors.
  1368. - Revert "Update to libusb-1.0.18"
  1369. - Disable fan/die clock control in hfa if the firmware does not support it, with
  1370. notification.
  1371. - Add ability to enter ANU frequency as a multiple of 25 from 150-500.
  1372. - Decrease hfa clock by 10 if a reset is attempted due to the device remaining
  1373. idle.
  1374. - ifdef out icarus options unused without icarus built in.
  1375. - Reorder command line options alphabetically.
  1376. - Add no matching work to hfa API output.
  1377. - Change various logging message levels in the hfa driver.
  1378. - Only adjust clocks if there is no restart in hfa to avoid 2 restarts back to
  1379. back.
  1380. - Ensure we iterate over all dies adjusting temperate for hfa by starting
  1381. iterating after the last die modified.
  1382. - Clamp initial hfa fanspeed to min/max if passed as parameters.
  1383. - Allow hfa fanspeed to be set via command line.
  1384. - Further relax the target temperatures on hfa driver, targetting 88 degrees.
  1385. - Try one more time to get the hfa header on init since it can take 2 seconds
  1386. for all 3 boards on a sierra.
  1387. - Update authors for removal of gpu/scrypt.
  1388. - Wait for 5 temperature updates in hfa before adjusting fanspeed.
  1389. - Have some leeway before starting to throttle hfa dies.
  1390. - Use increments of 10 when increasing hfa clock since it may not have 5 MHz
  1391. granularity internally.
  1392. - Only perform a hfa fan speed update if we have new temps to work with.
  1393. - Correctly measure the hfa max temp and smooth out the changes in its value.
  1394. - Choose better defaults for min/max/default fan settings for hfa driver.
  1395. - bab - reduce def speed, fix speed staying in ranges and report bank/chips in
  1396. ioctl() errors
  1397. - bab - add info about number of boards/chips to each Dead Chain
  1398. - These may not be longs (eg: OSX)... fo a safe cast to ensure.
  1399. - bab - add dead boards and dead chains to stats
  1400. - Add fanspeed to hfa api output and set initial fanspeed to 10%
  1401. - Add hfa fanspeed control to try and maintain a target temperature.
  1402. - API-README correct new text format documentation
  1403. - API allow multiple commands/replies in one request
  1404. - Add op commands necessary to control hfa fanspeeds.
  1405. - Add OP_FAN to hf protocol header.
  1406. - Always show the stratum share lag time in debug mode.
  1407. - Add stratum share response lag time to verbose output if it's greater than 1
  1408. second.
  1409. - Add stratum share submission lag time to verbose information if it's over 1
  1410. second.
  1411. - Check for more interrupted conditions in util.c and handle them gracefully.
  1412. - Send a ping to hfa devices if nothing is sent for over 5 seconds.
  1413. - Add OP_PING to hfa commands
  1414. - Display the hfa serial number as a hexadecimal value.
  1415. - Add the ability to display a hexadecimal 32 bit unsigned integer to the API.
  1416. - Limit all hfa restarts for temperature control to no closer than 15 seconds
  1417. apart.
  1418. - Allow the hfa temp target to be disabled by setting it to zero.
  1419. - Handle interruptions to various select calls in util.c
  1420. - Add sanity check for silly overflows in hfa die temperature readings.
  1421. - Add per-die throttling control for hfa driver based on each die's temperature,
  1422. issuing a suitable reset to maintain the temperature below a configurable target
  1423. temperature.
  1424. - Update hf protocol
  1425. - Do not memcpy in usbutils unless data was transferred.
  1426. - Send a full allotment of jobs to the hfa device after a restart instead of
  1427. reading the status.
  1428. - Export the flush_queue function for use by drivers.
  1429. - Remove wrong goto
  1430. - Remove the unqueued work reference when we discard work from get queued as
  1431. well.
  1432. - Wake the global work scheduler when we remove a work item from the unqueued
  1433. work pointer.
  1434. - Discard work that is stale in the get_queued() function, returning NULL
  1435. instead.
  1436. - Add a call to a driver specific zero stats function when zero stats is called
  1437. to allow each driver to reset its own stats as well if desired.
  1438. Version 3.12.0 - 29th January 2014
  1439. - Add support for AntminerU1 devices with the icarus driver.
  1440. - Add antminer U1 to comment in udev rules.
  1441. - Do away with usb resets entirely since we retry on both pipe and io errors now
  1442. and they're of dubious value.
  1443. - Retry on usb IO errors instead of faking success.
  1444. - Check that we've cleared the pipe error after a clear request, not the err
  1445. value which is unchanged.
  1446. - Update to libusb-1.0.18
  1447. - Change hfa overheat limit to 90 degrees.
  1448. - Relax timeout in hf get header to 500ms to match the usb timeout.
  1449. - Minion - check/clear interrupts for all chips
  1450. - Set info work to null after it is freed in nf1 after a restart to prevent
  1451. double free later.
  1452. - The second_run bool in libbitfury should be per device. Microoptimise its and
  1453. job_switched usage, removing the unused results array for NF1 devices.
  1454. - Fix displayed diff when solo mining at >2^32 diff.
  1455. - bab - stop stale work accumulating
  1456. - bab - set the default SPI speed back to 96000
  1457. Version 3.11.0 - 25th January 2014
  1458. - Add hashfast documentation to ASIC README
  1459. - Support the variable HFA naming throughout the driver notices.
  1460. - Set the global hfa hash clock rate to equal the lowest if we are lowering it
  1461. for a device reset since it may be re-hotplugged after failing reset.
  1462. - Decrease the hfa clock rate if it is overclocked and we have had to try
  1463. resetting it.
  1464. - Put a sanity check on the measured temperature in the hfa driver for obviously
  1465. wrong values.
  1466. - Avoid calling applog from within hfa statline before to avoid a deadlock.
  1467. - Add throttling control to hfa driver, configurable at command line, nominally
  1468. set to 85 degrees.
  1469. - Reset hfa device if no valid hashes are seen for 1 minute from the last work.
  1470. - Store when the last getwork was retrieved and display it in the API summary.
  1471. - bab - also report dead chip count screen
  1472. - Count share based hashes in the hfa driver with the device diff to get more
  1473. frequent updates.
  1474. - Only count 2/3 of the accumulated hashes on each pass through the hfa scan
  1475. work loop to smooth out displayed hashrate.
  1476. - bab add total history HW% to API stats
  1477. - Test valid nonces in the hashfast driver allowing us to check against the
  1478. target when trying to submit them.
  1479. - No point casting a double to a uint64
  1480. - Convert the hfa hashmeter to one based on successful share return and display
  1481. the raw and calculated hash totals in the API.
  1482. - bab - remove libbitfury dependency since it requires USB
  1483. - Add description to hfa hash clock command.
  1484. - Add hfa board temperatures to API output.
  1485. - Wait for up to 0.5 seconds in the hashfast scanwork loop if no jobs are
  1486. required.
  1487. - Label HFA devices as B or S when their configuration matches babyjet or
  1488. sierra.
  1489. - Fix libbitfury being compiled in always by mistake.
  1490. - bab - spelling
  1491. - Add bab-options
  1492. - bab - tune the chip speed based on error rates
  1493. - bab record/report spie and miso errors
  1494. - Win32 falsely comes up as big endian pulling in the wrong hf protocol header.
  1495. - Remove unused components in hashfast driver.
  1496. - Check in all usb communication places for hashfast driver that the device
  1497. still exists.
  1498. - Do not send a usb reset on a usb read pipe error.
  1499. - Don't replace usb pipe errors with the pipe reset return code.
  1500. - Updated hf protocol header.
  1501. - The search for extra nonce is not worth performing in the hashfast driver.
  1502. - Add core address to hfa parse nonce debugging.
  1503. - Retry sending a frame once if it has failed in hfa_send_frame
  1504. - Add extra hfa usb init errors.
  1505. - Quiet now unused variable warning in hfa detect.
  1506. - Remove unused variable.
  1507. - Add board temperature to hfa debug
  1508. - Make submit_tested_work return a bool about whether it meets the work target
  1509. or not.
  1510. - Provide a helper function for determining dev runtime and use it in the
  1511. hashmeters used.
  1512. - Look for hfa usb init header for 2 seconds, then resend the init twice more
  1513. before failing.
  1514. - Really only set up the hfa crc table once.
  1515. - Generically increase the queue if we are mining on a pool without local work
  1516. generation each time we run out of work.
  1517. - Change new block detection message since longpoll is rarely relevant today.
  1518. - Change the default clockspeed bits on nanofury devices to 50 and add a command
  1519. line option to allow it to be changed.
  1520. - Use unused line at the top of the log window which often gets stuck
  1521. unchanging.
  1522. - Clear pool work on a stratum reconnect message.
  1523. - bab record/report spie and miso errors
  1524. - bab - cleanup old work for dead chips also
  1525. - bab add avg fail tests to API stats
  1526. - bab report bank/board/chip for dead and v.slow chips
  1527. - bab process all nonce replies per chip together
  1528. - bab reduce work delays
  1529. - bab record the number of E0s discarded
  1530. - bab - modified result parsing
  1531. - bab restore removed unused flag
  1532. - configure - correct minion name
  1533. - bab only scan valid nonce offsets
  1534. - bab record continuous (and max) bad nonces
  1535. - bab display Banks/Boards/Chips in the device window
  1536. - Modify thread naming to make them easier to identify
  1537. - bab reduce the work send delay
  1538. - bab remove results polling
  1539. - bab report SPI wait in seconds
  1540. - bab report missing chips at start and API
  1541. - bab ensure there's enough space for the nonce reply
  1542. - bab correct stats 'Send Max'
  1543. - bab allow long enough wait on ioctl() per board
  1544. - bab more I/O stats
  1545. - api.c 2014
  1546. - api allow any size stats data
  1547. - bab add processed links which excludes expired links skipped
  1548. - bab report chips per bank, hw% and ghs per chip
  1549. - bab lock access to new_nonces to ensure correct reporting
  1550. - bab report V2 banks/boards during initialisation
  1551. - bab expire chip work
  1552. - bab use only k_lists and make work handling more refined
  1553. - klist - allow adding to tail
  1554. - bab remove old unused #define
  1555. - bab correct for master git
  1556. - correct klist reallocs
  1557. - klist lists for bab
  1558. - api.c correct DEVICECODE and ordering
  1559. - Maxchips should be 384 (16 chips/board 24 boards/controller)
  1560. - bab more detailed stats and delay less when waiting for a buffer
  1561. - api add data type AVG float 3 decimal
  1562. - bab - add V2 detect with bug fix in detect
  1563. - api.c set the actual version number to 3.0
  1564. - API V3.0 unlimited socket reply size
  1565. - README update --usb
  1566. - Check for loss of device in usb read before any other code on the usbdev
  1567. - Change stratum strings under stratum_lock in reconnect and free old strings.
  1568. - Add mcp2210 compilation to want_libbitfury configs.
  1569. - Fix HF driver typo.
  1570. Version 3.10.0 - 9th January 2014
  1571. - Set the mcp2210 transfer setting only when it changes.
  1572. - Buffer sizes in nanofury device data are unnecessarily large.
  1573. - Only perform spi reset on init, not with each transaction.
  1574. - Remove spi_detect_bitfury at nanofury startup and fix incorrect refresh time.
  1575. - Use a simple serialised work model for nanofury
  1576. - Use bitfury_checkresults to avoid hashing results twice in nanofury.
  1577. - Export bitfury_checkresults in libbitfury
  1578. - Pass extra parameters for later use in libbitfury_sendHashData
  1579. - Avoid double handling bswap of the nonce value in nanofury
  1580. - Avoid unnecessary rehashing in nanofury nonce checking.
  1581. - Remove the unused portions of atrvec in the nanofury driver
  1582. - Age work in nf1_scan to avoid risk of losing a work item and leaking memory.
  1583. - bitfury_work_to_payload is double handling the data unnecessarily
  1584. - Default bitrate on nanofury should be 200kHz
  1585. - localvec should be only 80 bytes not 80 words
  1586. - Wrong init value for nanofury
  1587. - Remove unused rehash values from nanofury driver.
  1588. - Only update info work in nanofury driver when it's empty.
  1589. - Fill the appropriate type of usb transfer when we know if it's an interrupt
  1590. transfer instead of a bulk one.
  1591. - Use the internal knowledge of the usb epinfo to determine whether we should be
  1592. doing an interrupt instead of a bulk transfer, and do not send a ZLP if so, and
  1593. limit read transfer to expected size automatically.
  1594. - Avoid bin2hex memleak when we start getting nanofury nonces
  1595. - Set atrvec only once and use a local array for each device's work.
  1596. - Cancel any spi transfers on nf1 close
  1597. - Add bitfury detection loop to nanofury startup
  1598. - Move spi init code to libbitfury
  1599. - Remove inappropriate extra config reg in nanofury setup.
  1600. - Status 0x30 should never happen with spi transfers.
  1601. - Fix spi transfer data size transmission mistakes.
  1602. - Minor correctness change in spi_add_data
  1603. - spi_txrx should always send and receive the same size message
  1604. - Random libbitfury changes.
  1605. - Set value of gpio pins to low on closing nanofury.
  1606. - Fix more init sequence for nanofury.
  1607. - Add basic initialisation for nf1 devices
  1608. - Add basic nf1_scan function.
  1609. - Basic import of libbitfury functions from nanofury branch
  1610. - Import functions from nanofury fork for libbitfury
  1611. - Meter out spi sends to only 2 bytes at a time, offsetting according to how
  1612. much data returns.
  1613. - Use the usb read limit function for mcp2210 reads.
  1614. - Provide a way for usb reads to just read the size asked for with a limit bool.
  1615. - Get pin value after an nf1 spi reset.
  1616. - Make sure what we send in the buffer doesn't change during spi reset for
  1617. nanofury
  1618. - Remove all standalone gpio setting change functions in mcp2210 and just use
  1619. the one global setting function.
  1620. - Set gpio values in the one function with all values for nanofury.
  1621. - Provide a helper function for setting all mcp2210 gpio settings.
  1622. - Add a helper function for getting all mcp2210 gpio settings.
  1623. - Set all pin designations and directions in one call for nanofury and don't
  1624. bother storing their values in the info struct.
  1625. - Provide helper functions for setting all pins and dirs on mcp2210
  1626. - Set all nanofury pin designations in one call
  1627. - Provide a helper function for setting all pin designations on mcp2210
  1628. - Store the spi settings in a struct for nanofury devices.
  1629. - Check the received status in mcp2210 spi transfers and repeat a zero byte send
  1630. if it's in progress.
  1631. - Set the bytes per spi transfer prior to each mcp2210 transfer.
  1632. - Separate out the send and receive functions for mcp2210 and check response
  1633. value in return.
  1634. - Check that mcp2210 spi settings have taken and check the value of the pin
  1635. during nanofury setup.
  1636. - Don't set GPIO pin designations after initial setting in nanofury since the
  1637. direction and values will be changed.
  1638. - Provide an mcp 2210 set gpio input helper function that sets a pin to gpio and
  1639. input.
  1640. - Move the set gpio output function to a generic mcp2210 version from nanofury
  1641. which also sets the pin to gpio.
  1642. - Implement a nanofury txrx with a larger buffer and cycling over data too large
  1643. to send.
  1644. - Implement magic spi reset sequence for nanofury.
  1645. - Add more spi magic to the nanofury init sequence.
  1646. - Add lots of magic spi initialisation to nanofury.
  1647. - Export reused components of bitfury management into a libbitfury and use for
  1648. bab and bitfury drivers.
  1649. - More init sequence for nanofury and implement a close function that sets all
  1650. pins to input.
  1651. - Reword offset header handling in hfa_get_header
  1652. - Sanity check in hfa_get_header
  1653. - Add more checks in hashfast driver for lost devices.
  1654. - Change spimode and send more data in nanofury setup.
  1655. - Add basic setup comms to nanofury.
  1656. - Implement an mcp2210 spi transfer function.
  1657. - Set the initial spi settings for nanofury driver.
  1658. - Provide a helper function for gettings mcp2210 spi settings.
  1659. - Implement an mcp2210 set spi transfer settings function.
  1660. - Cancel any SPI transfers in progress in nanofury after initial setup.
  1661. - Implement an mcp2210 spi cancel function.
  1662. - Return only binary values for mcp2210 GPIO values.
  1663. - Set GPIO LED and power to high in nanofury driver.
  1664. - Implement initial part of nanofury init sequence for GPIO pin settings and add
  1665. output debugging of set values.
  1666. - Add helper functions for getting and setting mcp2210 gpio pin designations.
  1667. - Don't return an error in usb read if we've managed to get the whole read
  1668. length we've asked for.
  1669. - Use correct endpoint order for nanofury devices and read with a short timeout
  1670. on return loop from send_recv.
  1671. - Add mcp2210 helper functions for getting and setting one GPIO pin val and
  1672. direction.
  1673. - Create a generic gpio pin struct and add helpers for mcp get pin val and dirs.
  1674. - Check the receive msg of a send/receive cycle on mcp2210 matches the send
  1675. message.
  1676. - Add a set of usb commands to the usbutils defines for mcp2210 comms, and use
  1677. the same command name for send and receive.
  1678. - Create a generic mcp2210 send_rcv function.
  1679. - Include mcp header for bitfury and fix extra params in macro.
  1680. - Add basic SPI comms defines for mcp2210 and build rules for bitfury.
  1681. - Minion set some core defaults similar to final requirements
  1682. - minion compile warnings
  1683. - move driver-minion.c to main directory
  1684. - Minion with ioctl() stats, settings to attempt to emulate 21TH/s
  1685. - minion driver with results interrupt working
  1686. - tested working driver-minion.c without interrupts
  1687. - Working driver-minion.c v0.1
  1688. - driver-minion.c compilable untested
  1689. - minion driver - incomplete
  1690. - Add minion driver into cgminer
  1691. - Add basic device detection and updated udev rules for nanofury devices.
  1692. - Remove GPU from share logging example.
  1693. - Don't keep resetting BXF clockspeed to default.
  1694. - If no pools are active on startup wait 60s before trying to reconnect since we
  1695. likely have the wrong credentials rather than all the pools being out.
  1696. - Discard bad crc packets for hashfast driver instead of trying to process them.
  1697. - Update documentation for modified avalon options syntax and document relevant
  1698. 55nm details.
  1699. - Modify the auto tuning sequence to work with the 50MHz changes required to
  1700. work with 55nm Avalon.
  1701. - 55nm avalon requires the delays between writes reinstated for stability.
  1702. - Use an equation instead of a lookup table to set the frequency for 55nm avalon
  1703. allowing arbitrary values to be used.
  1704. - Make the result return rate low detection on avalon less trigger happy.
  1705. - Always send the bxf device a clockspeed after parsing the temperature in case
  1706. the device has changed the clockspeed itself without notification.
  1707. - Fix BXF being inappropriately dependent on drillbit.
  1708. Version 3.9.0 - 23rd December 2013
  1709. - drillbit asic - enable in api.c
  1710. - Fix trivial warnings in knc driver.
  1711. - Reinstate work utility based hashmeter for knc.
  1712. - drillbit format %z not valid on windows
  1713. - drillbit more formatting changes
  1714. - usbutils remove old code added back
  1715. - Memset the spi tx buffer under lock in knc driver.
  1716. - drillbit fix temp display to fit in standard space
  1717. - Drillbit formatting
  1718. - drillbit - use one drvlog and display dname before add_cgpu
  1719. - Keep orginal naming for the bitfury driver
  1720. - knc: Bugfix - good shares wrongly reported as HW errors. Root cause of the
  1721. problem: several work items were assigned the same work_id in the active works
  1722. queue of the knc driver. Thus when good nonce report arrived from the FPGA,
  1723. wrong work item was picked up from the queue, and submit_nonce evaluated that
  1724. as an error. Fix: Limit the work_id counter update rate. Update it only to the
  1725. number of works actually consumed by the FPGA, not to the number of works
  1726. send.
  1727. - Store per-chip submit information for bxf device and show them in the API.
  1728. - Check for removed bxf devices before trying to update work or send messages.
  1729. - api.c no decref if not json
  1730. - Minimise risk of nonce2 overflow with small nonce2 lengths by always encoding
  1731. the work little endian, and increasing the maximum size of nonce2 to 8 bytes.
  1732. - Change default hashfast timeout to 500ms.
  1733. - Ensure we can look up the work item in the hashfast driver or print out an
  1734. error if we don't.
  1735. - Drillbit source formatting - reindent and retabify
  1736. - Add ASIC count, temperature status to drillbit API output (closes #1)
  1737. - Many warning fixes
  1738. - knc: Do not include variable "last minute" data into the "last hour" per-core
  1739. stats
  1740. - knc: Make per-core statistics available through API
  1741. - Implement command line control of the bxf target temperature.
  1742. - Add a simple PID-like controller to bi*fury devices to dynamically alter the
  1743. clock setting to maintain a nominal target temperature set to 82 degrees.
  1744. - Add data to BXF API output.
  1745. - Add support for newer protocol bi*fury commands job, clock and hwerror,
  1746. setting clock to default 54 value, turning parsing into a compact macro.
  1747. - Look for the thermal overload flag in the gwq status message in the hashfast
  1748. driver and send it a shutdown followed by an attempted reset.
  1749. - Log message fixups
  1750. - Fix for "Timing out unresponsive ASIC" for pools which send early reconnect
  1751. requests, and then take a short time to send work (ie BTCGuild)
  1752. - Shorten initial config line, win32/pdcurses doesn't like long lines during
  1753. early logging
  1754. - Pull back the very long timeouts set in fe478953cf50
  1755. - Fix bug where work restart during results scan could lead to bad device state
  1756. - Align device status lines same regardless of number of temp status or >10
  1757. ASICs
  1758. - Tag log lines from brand new devices as DRB-1 until they are initialised
  1759. - Tag log lines as 'DRB0' rather than 'DRB 0', same as other places in cgminer
  1760. - Print a summary of the device settings at level NOTICE during initialisation
  1761. - Allow chosing device settings based on 'short' product names shown in status
  1762. line
  1763. - Allow per-device settings to use "DRBnn" as an identifier instead
  1764. - Issue an ASIC restart during a work_restart, removes spurious timeout messages
  1765. from ASICs and probably some rejected shares
  1766. - Check all results against all work instead of just taking the first match
  1767. (avoids some rejected submissions to the pool, ASIC can produce multiple
  1768. candidate results.)
  1769. - Fix memory leak caused by unnecesarily copied work
  1770. - Fix bug with find_settings not returning default value
  1771. - Set timeouts on write, set very long timeouts
  1772. - Merge drillbit driver
  1773. Version 3.8.5 - 10th December 2013
  1774. - Increase the BFLSC overtemp to 75 for fanspeed to maximum.
  1775. - Set bflsc cutoff temperature to 85 degrees and throttle 3 degrees below the
  1776. cutoff temperature.
  1777. - Only set LIBUSB_TRANSFER_ADD_ZERO_PACKET for libusb versions we know include
  1778. support for.
  1779. - Provide a helper function that can reset cgsems to zero.
  1780. - Add to cgminer_CPPFLAGS instead of redefining them.
  1781. - Attempt a libusb reset device on usb devices that have stopped responding.
  1782. - Replace deprecated use of INCLUDES with _CPPFLAGS.
  1783. - Remove more unused GPU code.
  1784. - Attempt USB device resets on usb read/write errors that will normally cause
  1785. the device to drop out.
  1786. - Quieten down jansson component of build.
  1787. - Cache the bool value for usb1.1 in _usb_write
  1788. - Initialise usb locks within usbutils.c instead of exporting them.
  1789. - Imitate a transaction translator for all usb1.1 device writes to compensate
  1790. for variable quality hubs and operating system support.
  1791. - Rationalise variables passed to usb_bulk_transfer.
  1792. - Unlink files opened as semaphores on releasing them.
  1793. - Remove user configuration flag from pll bypass enabling in hashfast driver.
  1794. - Provide an hfa-dfu-boot option for resetting hashfast devices for
  1795. reprogramming.
  1796. - Fixed one byte stack overflow in mcast recvfrom.
  1797. - Having changed C_MAX means we don't calloc enough for usb stats, off by one.
  1798. - Don't free the info struct on hashfast shutdown since it's still accessed
  1799. after a device is removed.
  1800. Version 3.8.4 - 1st December 2013
  1801. - Deprecate the usb usecps function and just split up transfers equal to the
  1802. maxpacketsize on usb1.1 devices.
  1803. - Retry sending after successfully clearing a pipe error.
  1804. - Drop logging of timeout overrun message to verbose level.
  1805. - Use a much longer callback timeout for USB writes on windows only as a last
  1806. resort since cancellations work so poorly.
  1807. - Use vcc2 in bflsc voltage displayed.
  1808. - Increment per core errors on false nonces in bflsc and add per core statistics
  1809. to api stats, removing debugging.
  1810. - Store a per-core nonce and hw error count for bflsc.
  1811. - Fix json parsing in api.c
  1812. - Add debugging to hfa driver for how many jobs are being sent.
  1813. - Shut down the hfa read thread if the device disappears.
  1814. - Add debug output saying what frame command is being sent in hfa driver.
  1815. - Revert "Disable USB stats which were not meant to be enabled by default and
  1816. add extra memory for a memory error when stats are enabled."
  1817. - Reset work restart flag in hfa driver since we may check for it again in
  1818. restart_wait.
  1819. - Add more op usb init errors for hfa driver.
  1820. - Perform basic displaying of hfa notices received.
  1821. - Add hfa op usb notice macros.
  1822. - Update hf protocol header.
  1823. - Use sync usb transfers in lowmem mode.
  1824. - Go back to allowing timeout errors on USB writes to be passed back to the
  1825. driver without removing the device in case the driver wishes to manage them.
  1826. - Initialise more values for the hfa data structures.
  1827. - A USB control error must be < 0
  1828. - Simplify USB NODEV error checking to success only for writes and control
  1829. transfers, and success and timeout for reads.
  1830. - libusb error IO should be fatal as well if it gets through usb read and write.
  1831. - Allow IO errors in usb reads/writes to be ignored up to retry max times.
  1832. - Use correct padding for bxf temperature display.
  1833. - Initialise devices before attempting to connect to pools to allow their thread
  1834. prepare function to be called before having to connect to pools.
  1835. - Add hidden hfa options to set hash clock, group ntime roll and pll bypass,
  1836. fixing frame sent on reset to include extra data.
  1837. - Relax the timeouts for the slower usb devices on linux.
  1838. - Add big endian hf protocol header to Makefile
  1839. - Check for correct big endian macro in hf_protocol
  1840. - Use an absolute timeout in hfa_get_header to cope with buffered usb reads
  1841. returning instantly confusing the 200ms counter.
  1842. - Update hfa_detect_one to use the new detect function API.
  1843. Version 3.8.3 - 23rd November 2013
  1844. - Set the bitfury device start times from when we first get valid work.
  1845. - Fix stack corruption of zeroing too much in bf1 driver.
  1846. - Make usb_detect return the cgpu associated with it to check if it succeeds to
  1847. decide on whether to increment the device count or not.
  1848. - Set tv work start time for bxf driver.
  1849. - Age the bxf work items over 90 seconds, not the bf1 work items.
  1850. - Zero the read buffer in _usb_read to avoid stale data and only use stack
  1851. memory instead of using the bulkbuf since it is only used in _usb_read.
  1852. - Leave room for temperatures above 100 degrees and pad consistently for bxf
  1853. statline.
  1854. - Drop json stratum auth failed message log level to verbose.
  1855. - Change the processed value not the bufsiz in response to an end of message
  1856. marker.
  1857. - Don't lose data beyond the end of message in a usb read.
  1858. - Silence irrelevant warning.
  1859. - Only check strlen on end if end exists.
  1860. - Simplify the end of message detection in _usb_read and allow it to return
  1861. without doing another read if the message is already in the buffer.
  1862. - Increase work ageing time to 90 seconds for bxf driver to account for firmware
  1863. changes.
  1864. - Use the age_queued_work function in the bitfury driver.
  1865. - Provide a function to discard queued work based on age.
  1866. - The json_val in api.c is a borrowed reference, not a new one so don't decref
  1867. it.
  1868. - Decrement json references in api.c to not leak memory.
  1869. - line 2913 added urlencode
  1870. - With reliable writes to the avalon there is no need for the sleep delays
  1871. between writes.
  1872. - There is no need to limit usb write transfers to maxpacketsize and it's
  1873. harmful for large transfers on slow devices such as wrt routers.
  1874. - Disable USB stats which were not meant to be enabled by default and add extra
  1875. memory for a memory error when stats are enabled.
  1876. - Set limit and count to integers to not overflow during failed hotplug attempts
  1877. and then not trying again.
  1878. - Update api example compilation instructions.
  1879. Version 3.8.2 - 16th November 2013
  1880. - Add more verbose documentation to the readme files for windows users.
  1881. - Add more information on libusb failure to init telling users to check README
  1882. file.
  1883. - Add information on unloading cdc drivers on osx to README
  1884. - Prevent a deadlock with use of restart_threads by spawning a thread to send
  1885. the driver flush work messages.
  1886. - Set priority of various threads if possible.
  1887. - Add bxf data to api output.
  1888. - Do not hold the mining thread lock in restart_threads when calling the driver
  1889. flush work commands.
  1890. - Send extra work regularly to the bxf device and parse the needwork command by
  1891. sending the amount of work it requests.
  1892. - Allow messages to have arbitrary offsets in the bxf parser in case we have
  1893. lingering buffered data.
  1894. - Send the maxroll command to the bxf driver and store the value to see if we
  1895. need to update it.
  1896. - Add sending of flush command to bxf on flush_work
  1897. - Add flush and version commands to bxf start up, flush buffer and try to parse
  1898. version response string.
  1899. - Abstract out bxf recv message.
  1900. - Add extra bxf commands to usbutils
  1901. - Abstract out bxf send message to allow us to easily add extra commands.
  1902. - Don't run device restart code if the device is not enabled.
  1903. - Expand size of bitfury statline
  1904. - Various driver fixes for bitfury devices, including a flag from when first
  1905. valid work appears.
  1906. - Look up work results in bxf driver from correct variable.
  1907. - Correct incorrect error code in bxf driver for usb writes and add debugging.
  1908. - Add bxf details to usbutils.
  1909. - Implement a statline showing temperature for bxf
  1910. - Add api data for bxf device, sharing the hashrate function with bf1.
  1911. - Count no matching work as a hw error on bxf
  1912. - Add BXF to udev rules.
  1913. - Work id should be hexadecimal in bxf messages.
  1914. - Add unrecognised string debugging to bxf driver.
  1915. - Implement the main scanloop for bxf, trying to prevent it from ntime rolling
  1916. work if the work protocol does not allow it.
  1917. - Parse bxf work submits fully, submitting the results.
  1918. - Provide a function for setting the work ntime.
  1919. - Implement a skeleton parse bxf submit function.
  1920. - Use the bxf read thread to set the device target and send its first work item.
  1921. - Implement a bxf send work function and set update and restart functions to
  1922. sending new work since that's the equivalent for that device.
  1923. - Add temperature parsing to bxf driver
  1924. - Create and destroy a basic bxf read thread.
  1925. - Remove the buffer from bitfury info since it is only used on one pass in the
  1926. bf1 device.
  1927. - Add a rudimentary bxf detect one function.
  1928. - Rename all bf1 specific functions in the bitfury driver, using a switch to
  1929. choose correct function.
  1930. - Rename bitfury_getinfo to bf1_getinfo since it's unique to bf1 devices.
  1931. - Separate out the bf1 reset from bitfury reset.
  1932. - Store the bitfury identity in the info struct.
  1933. - BaB - updated tested OS comment
  1934. - Uniquely identify the BF1 and BXF bitfury devices.
  1935. - Remove the default libusb WinUsb pipe policies that don't suit us.
  1936. - Only set the winusb pipe policy if it doesn't match our requirements instead
  1937. of every transfer.
  1938. - klondike - dont try to flush if not initialised
  1939. - api.c trylock() add missing locklock
  1940. - Use our new zero length packet support directly in windows.
  1941. - Enable support for zero length packet on windows and auto clear pipe stalls.
  1942. - util.c: Decreasing reference count on allocated JSON obects to prevent memory
  1943. leak
  1944. - api.c: Release apisock on error in api()
  1945. - api.c: Release io_data->ptr when releasing io_data in io_free()
  1946. - We can't connect to a GBT pool at all with fix protocol enabled.
  1947. - Initialise the stgd lock mutex earlier to prevent dereferences when pool
  1948. testing occurs before it.
  1949. - Klondike support I2C USB layout also - as KLI
  1950. - Return error codes in avalon_read() if they're not timeouts.
  1951. - Break out of the avalon idle loop if we get a send error.
  1952. - Set avalon ftdi latency to just less than the time it would take to fill the
  1953. ftdi buffer at 115200 baud
  1954. - Update example.conf
  1955. - Only limit packetsize on usb out writes.
  1956. - We must chop up every 64 bytes returned on an ftdi chip, not just the first 2
  1957. bytes so revert to parsing the data internally in the avalon instead of using
  1958. usbutils' simple ftdi parser.
  1959. - Only retry 3 times in hfa_reset.
  1960. - Only add_cgpu in hashfast driver once we have a real driver set up.
  1961. - Clean up properly if hfa_detect_common fails in the hashfast driver.
  1962. - --shares should be scaled to diff1 not absolute number of shares
  1963. Version 3.8.1 - 11th November 2013
  1964. - Revert "Send a zero length packet at the end of every usb transfer on windows
  1965. in case libusb internally has batched them into one maxpacket sized."
  1966. Version 3.8.0 - 10th November 2013
  1967. - api update version to 2.0 and remove GPU form API-README
  1968. -Remove now unused scrypt files.
  1969. - api.c remove all GPU/gpu references and correct code as required
  1970. - Rudimentary removal of GPU OpenCL and Scrypt features from api.c
  1971. - Reorder configure alphabetically for devices to compile and fail if no support
  1972. is selected to be compiled in.
  1973. - BaB update/format some comments
  1974. - BlackArrowBitfury early GPIO V1 driver
  1975. - Fine tune the reading of results in bitfury driver to not lose any across work
  1976. restarts or corrupt due to store results not parsed during restart.
  1977. - Send a zero length packet at the end of every usb transfer on windows in case
  1978. libusb internally has batched them into one maxpacket sized.
  1979. - Framework for ntime rolling, keep looking for OP_USB_INIT replies when other
  1980. packets received
  1981. - Configure source for a new BaB driver
  1982. - sha2 allow external access to some macros and the K array
  1983. - Fixed a math issue when reporting fan speed on the status line.
  1984. - Use the main hashlist to store work done in the bitfury driver and remove work
  1985. from the list by time, thereby fixing the duplicates at startup. Count hardware
  1986. errors for when no match occurs.
  1987. - Add a get and queue helper work function.
  1988. - Remove GPU mining code.
  1989. - Use libusb's own zero length packet support unless we have to emulate it on
  1990. windows since only libusb knows for sure if it's needed.
  1991. - Unlock the avalon qlock while sending tasks to not hold the lock for an
  1992. extended period.
  1993. - Sleep in avalon send task on return to the function to allow other code to
  1994. work during the sleep period.
  1995. - Send zero length packets when terminating a usb write aligned to
  1996. maxpacketsize.
  1997. - Do the driver flush in avalon code lockless since it can lead to deadlocks.
  1998. - Reset the work_restart bool after the scanwork loop in case the driver flushes
  1999. work synchronously.
  2000. - Only check for the stratum clean message if we have had a valid message.
  2001. - Get rid of the stage thread since all work can be asynchronously added now via
  2002. hash_push anyway.
  2003. - Remove the now incorrect faq entry regarding scrypt difficulty.
  2004. - Check for fatal read errors and break out of the read loop in avalon.
  2005. - Send errors are basically fatal in avalon driver so break out of the send
  2006. tasks loop.
  2007. - Make the avalon driver return -1 for hash count when usb fails, allowing the
  2008. main loop code to send it the shutdown flag.
  2009. - Break out of the hash work loops when a failure is detected instead of
  2010. dropping into mt disable.
  2011. - Use usbutils' own ftdi parser for avalon and the ftdir's own latency for
  2012. managing timeouts since we can wait on reads with completely asynchronous
  2013. reads+writes.
  2014. - Use usbutils' own cps function for slowing rate of usb writes on avalon.
  2015. - Fix build for no libcurl
  2016. - Check length before submitting sync transfers
  2017. Version 3.7.2 - 5th November 2013
  2018. - Clean up completely on avalon shutdown.
  2019. - Use cgsem timed waits in avalon driver to not miss any queued wake ups to
  2020. account for async messages coming during a flush work.
  2021. - Statline before is too long on icarus that doesn't have monitoring.
  2022. - Different windows+usb combinations respond with varying levels of reliability
  2023. wrt timeouts so use a nominal extra 40ms before cancelling transfers that fail
  2024. to time out on their own.
  2025. - Do all hotplug_process under the write mining_thr_lock
  2026. - Fix for opt_worktime on big endian machines.
  2027. - Correct set_blockdiff for big endian machines.
  2028. - Make sure cgpu exists in the restart threads loop in cases of hotplug etc.
  2029. - Treat usb write timeout errors as unrecoverable.
  2030. - Transfer errors are filtered out in usbutils now so no need to look for them
  2031. in NODEV checks.
  2032. - Remove now unused entries from struct cg_usb_device
  2033. - Do not double up with checking for end of timeout measurements in usb
  2034. read/write.
  2035. - Do get_work in fill_queue without holding other locks.
  2036. - Initialise usb after all the locks and conditionals are initialised.
  2037. - Use only a trylock in flush queue to prevent deadlocks.
  2038. - Add a wr_trylock wrapper for pthread rw lock write trylock.
  2039. - Scale diff for scrypt when testing for block solves.
  2040. - Fix for non curses build.
  2041. Version 3.7.0 - 4th November 2013
  2042. - Use WRITEIOERR macro check for all usb writes.
  2043. - Always use a usb read buffer instead of having to explicitly enable it.
  2044. - Force unlocking of the console lock on restart to avoid corrupting the console
  2045. state when we finally quit.
  2046. - Never wait indefinitely for a pthread conditional in the hash_pop loop in case
  2047. the work scheduler misses the last wakeup.
  2048. - Make hash_pop signal the work scheduler each time it waits on the conditional
  2049. that it should look for more work.
  2050. - Discriminate between libusb transfer errors and regular libusb errors and make
  2051. sure to capture them all.
  2052. - Always read a full sized transfer for bulk reads.
  2053. - Deprecate preferred packet size functions in usbutils since they're unhelpful.
  2054. - Copy known transferred amount back to buffer for usb reads instead of
  2055. requested length.
  2056. - Treat timeout errors on usb writes as IO errors.
  2057. - Ignore iManufacturer from bitfury devices to support bluefury as well as
  2058. redfury.
  2059. - Add more debugging info for when usb details don't match.
  2060. - Look for timeout overruns in usb read/write.
  2061. - Use an int for usb_read/write to identify overruns.
  2062. - Use the callback timeout as a safety mechanism only on windows.
  2063. - Instead of using complicated sleeps to emulate characters per second on usb
  2064. writes, submit only as many characters as can be transferred per usb poll of
  2065. 1ms, and use timeouts in bulk transfers, cancelling transfers only as a
  2066. failsafe.
  2067. - Remove discarded work from quota used.
  2068. - Display works completed in summary and API data.
  2069. - Store how many work items are worked on per pool.
  2070. - Make each pool store its on reference for what the most current block is and
  2071. fine tune management of block change in shared pool failover strategies using
  2072. the information.
  2073. - Rationalise use of current_hash to a single hex string the length of the
  2074. previous block and display only the first non zero hex chars of the block in the
  2075. status window.
  2076. - Update uthash to latest.
  2077. - show_hash doesn't know the size of the string so hard code the max size.
  2078. - Remove as many initial zeroes as exist on share display, abstracting out a
  2079. hash show function to use across different submission mechanisms.
  2080. - Add missing endian swap functions for 64bits.
  2081. - Sanity check for absurd target setting and divide by zero.
  2082. - Abstract out conversion of a 256 bit endian number to a double, correcting
  2083. errors and use it for determining any magnitude share diff.
  2084. - Avoid the extra generation of a byte flipped hash2 in struct work and directly
  2085. use the LE work hash.
  2086. - Add a sanity check to avoid divide by zero crashes in set_target
  2087. - Calculate diff from target accurately for all 256 bits.
  2088. - Set a true 256bit binary target based on any diff value in set_target()
  2089. - Provide a copy_work_noffset function for copying a work struct but changing
  2090. its ntime.
  2091. - Make calls to flush queue and flush work asynchronous wrt to the main work
  2092. loops.
  2093. - Share is also above target for submit noffset nonce.
  2094. - Use round for displaying current pool diff.
  2095. - Use round for stratum share diff display instead of floor.
  2096. - Use round instead of floor for displayed pool difficulty.
  2097. - Allow arbitrary diffs to be tested against nonces via a test_nonce_diff
  2098. function.
  2099. - Abstract out the rebuilding of hash2 in work.
  2100. - Share is above, not below target, when it doesn't meet it.
  2101. - Add the ability to add uint8 and uint16 entities to api data.
  2102. - Use a non blocking connect with a 1 second select timeout when initiating
  2103. stratum to allow us to iterate over all IPs returned by getaddrinfo in round
  2104. robin DNS pools.
  2105. - Minor style changes to output.
  2106. - Revert two different hash_sequence(_head)'s to one variable, use
  2107. HF_SEQUENCE_DISTANCE in both places
  2108. - Remove duplicate HF_SEQUENCE_DISTANCE() macro, and duplicate hash_sequence
  2109. from info structure
  2110. - Change SEQUENCE_DISTANCE() macro to HF_SEQUENCE_DISTANCE()
  2111. - Structure changes for OP_NONCE, add big endian header
  2112. - klondike - initialise stat_lock
  2113. - klondike - better to unlock locks than to lock them twice :)
  2114. - Add copyright notice to knc driver.
  2115. - Trivial style changes to knc driver.
  2116. - Improve performance of work generation by optimizing hex2bin and bin2hex
  2117. - klondike - change options to clock and temptarget only
  2118. - klondike - fix another uninit dev warning
  2119. - klondike - downgrade 'late update' but add an idle detect - and correct error
  2120. levels
  2121. - klondike - fix isc uninit warning
  2122. - Use a mutex to protect data in the knc structure, to prevent loading more work
  2123. during a flush, and unlock and return to main between calls to get_queued_work.
  2124. - Use the existing device_data for knc state data.
  2125. - Only count successful nonces as hashrate in the knc driver.
  2126. - Fix trivial warnings in knc driver.
  2127. - Add KNC to api
  2128. - klondike - drop the device for hotplug if it's unresponsive
  2129. - usbutils - usb_nodev() allow a driver to drop a device
  2130. - klondike - single 'shutdown' and ensure it happens
  2131. - klondike remove SCNu8 - unsupported on windows
  2132. - Correctly calculate sleep_estimate in usbutils that may have been preventing
  2133. usecps from working.
  2134. - Use a sanity check on timeout on windows.
  2135. - Better HW error count; disable permanently those cores which fail often
  2136. - KnC driver: knc-spi-fpga ASIC driver
  2137. - Fixup jansson & libusb include paths when using separate build directory
  2138. - 'llround' is more suitable here than 'roundl'
  2139. - Silence warning if MAX/MIN is already defined
  2140. - Remove prebuild ccan/opt dependencies
  2141. - Reinstate block solve testing.
  2142. - Dramatically simplify the calculation of blockdiff.
  2143. - Simplify the set_target function, allowing it to work properly for fractional
  2144. diffs.
  2145. - Merge hashfast driver
  2146. - Merge KnC driver
  2147. Version 3.6.6 - 26th October 2013
  2148. - Remove inappropriate extra locking in _usb_transfer_read
  2149. Version 3.6.5 - 26th October 2013
  2150. - klondike - fix uninitialised dev bug
  2151. - Adjust the binary ntime data in submit_noffset_nonce even when there is no hex
  2152. ntime string for eg. gbt.
  2153. - Put an entry into the work struct telling drivers how much they can roll the
  2154. ntime themselves.
  2155. - Only set libusb cancellable status if the transfer succeeds.
  2156. - Remove the applog on miner threads dying to prevent deadlocks on exit.
  2157. - Do one extra guaranteed libusb event handling before testing if there are any
  2158. pending async usb transfers.
  2159. - Use a linked list for all usb transfers instead of just cancellable ones.
  2160. - Provide a mechanism for informing drivers of updated work templates for
  2161. stratum and gbt mining.
  2162. - Add cancellable transfers correctly to the ct_list
  2163. - Check for presence of thr in icarus get nonce for startup nonce testing to
  2164. work.
  2165. - Use cancellable usb transfers in the icarus driver to avoid having to loop and
  2166. poll when waiting for a response and to speed up work restart response time.
  2167. - Add a usb_read_ii_timeout_cancellable wrapper
  2168. - Add usb transfer cancellation on shutdown and documentation regarding where
  2169. cancellable transfers are suitable.
  2170. - Use cancellable transfers on bitfury device.
  2171. - Cancel cancellable usb transfers on work restart messages.
  2172. - Don't bother having a separate cancellable transfer struct for usb transfers,
  2173. simply include the list in the usb_transfer struct.
  2174. - Add wrappers for usb_read_cancellable and usb_read_timeout_cancellable
  2175. - Specifically set the cancellable state for it to not be uninitialised in the
  2176. usb transfer struct.
  2177. - Alter the usb cancellable list only under cgusb_fd_lock write lock.
  2178. - Pass the cancellable option to _usb_read options to decide on whether to add
  2179. usb transfers to the list of cancellable transfers.
  2180. - Create a linked list of potentially cancellable usb transfers.
  2181. - Don't attempt to disable curses or print a summary during an app restart to
  2182. prevent deadlocks.
  2183. - Keep the libusb event handle polling thread active until there are no async
  2184. usb transfers in progress.
  2185. - Keep a global counter of how many async usb transfers are in place.
  2186. - Perform libusb_submit_transfer under the write variant of cgusb_fd_lock
  2187. - klondike - error condition handling
  2188. - Avoid entering static libusb directory if --with-system-libusb is enabled.
  2189. - Minor opencl build corrections.
  2190. - Enable dynamic linking against system libusb --with-system-libusb
  2191. - Modify Makefile to only include opencl related code when configured in.
  2192. - Convert opencl to need to be explicitly enabled during build with
  2193. --enable-opencl
  2194. - Implement a cglock_destroy function.
  2195. - Implement a rwlock_destroy function.
  2196. - Implement a mutex_destroy function.
  2197. - Add usb command name to critical libusb error reporting.
  2198. - Use windows' own higher resolution time and handlers allowing us to have
  2199. higher precision absolute timeouts.
  2200. - Fix lldiv error in windows cgminer_t calculation.
  2201. - miner.php correct sort gen field names largest to smallest
  2202. - api ... the code related to device elapsed
  2203. - api add device elapsed since hotplug devices Elapsed is less than cgminer
  2204. Elapsed
  2205. - Drop usb buffering message to debug logging level.
  2206. - Do the ntime binary modification to the work struct when submitting an ntime
  2207. offset nonce within submit_noffset_nonce
  2208. - Code cleanup and improved documentation
  2209. - Improvements to support for BitBurner boards
  2210. - Convert libusb transfer errors to regular libusb error messages to allow for
  2211. accurate message reporting.
  2212. Version 3.6.4 - 18th October 2013
  2213. - Fixing the memory leak for remaining semaphores means we can go back to using
  2214. async transfers on other OSes with our own timeout management again.
  2215. - Use the forcelog function on shutdown to cope with indeterminate console lock
  2216. states due to killing of threads.
  2217. - Add a forcelog variant of applog which invalidates any console lock to force
  2218. output.
  2219. - Send pthread_cancel to failed completion_timeout that has timed out.
  2220. - Simplify queued hashtable by storing unqueued work separately in a single
  2221. pointer.
  2222. - bflsc use getinfo chip parallelization if it is present
  2223. - bflsc - fix brackets so [Chips] isn't always null
  2224. - Remove unused variables.
  2225. - Use cgcompletion timeouts for the unreliable shutdown functions on kill_work.
  2226. - Fix cgcompletion return code and free on successful completion.
  2227. - Provide a cg_completion_timeout helper function for unreliable functions that
  2228. takes arbitrary functions and parameters and reliably returns.
  2229. - Perform sync transfers on shutdown to allow final transfers to complete.
  2230. - Destroy cgsems used after transfers to not leave open files on osx.
  2231. - klondike rewrite work control
  2232. - allow __work_complete() access
  2233. - miner.h allow devices to tv_stamp work
  2234. Version 3.6.3 - 17th October 2013
  2235. - API add 'MHS %ds' to 'summary'
  2236. - Optional lock tracking and stats via the API
  2237. - Speed up polling repeat again in usb poll thread and handle async after the
  2238. message to disable polling is complete.
  2239. - Revert to using timeouts on !linux since libusb leaks memory without them.
  2240. - Revert to libusb instead of libusbx
  2241. Version 3.6.2 - 17th October 2013
  2242. - Remove unused components of jansson
  2243. - Remove unused parts of libusb
  2244. - Work around older libtoolize that fails without top ltmain.sh not being
  2245. present during autogen
  2246. - Fix open coded use of autoreconf in autogen
  2247. - Update jansson to only build parts we require and suited to our build
  2248. environment.
  2249. - Initial import of jansson-2.5
  2250. - Prevent further USB transfers from occurring once the shutdown signal has been
  2251. sent to prevent transfers getting stuck and libusb failing to shut down.
  2252. - Make the USB polling thread poll every second to potentially aid longer
  2253. timeout transfers.
  2254. - Set device_diff on work in get_work to not be missed with drivers that use
  2255. get_work directly.
  2256. - Convert icarus driver to hash_driver_work model.
  2257. - bflsc - also allow ' 0' in DEVICES IN CHAIN
  2258. - bflsc - allow a 0 in DEVICES IN CHAIN
  2259. - Add needed EXTRA_DIST for libusbx.
  2260. - Update libusbx configure.ac changes.
  2261. - Revert libusb Makefile changes from going to libusbx.
  2262. - Fix trivial libusbx warnings.
  2263. - Convert libusb-1.0.16-rc10 to libusbx-1.0.17
  2264. Version 3.6.1 - 14th October 2013
  2265. - Emulate the libusb_control_transfer sync setup in our async variant.
  2266. - usbutils - make all libusb_error_name messages the same
  2267. Version 3.6.0 - 14th October 2013
  2268. - increasing max miners for avalon driver
  2269. - using separate identifier for bitburner fury boards
  2270. - changes to bitburner driver for bitburner fury boards
  2271. - hexstr is too small in test_work_current
  2272. - Windows uses errno for WSAETIMEDOUT
  2273. - Convert the usb callback function to using cgsem_t timed waits to avoid race
  2274. conditions with conditionals/mutexes.
  2275. - Give correct return code in cgsem_mswait
  2276. - Check for correct timeout error in cgsem_mswait
  2277. - Fix util.h exports for cgsem_mswait
  2278. - Implement a generic cgsem_mswait similar to sem_timedwait
  2279. - Use the one LIBUSB_ERROR_TIMEOUT for cancelled transactions since this error
  2280. is explicitly tested for in various drivers.
  2281. - Do not use locking on usb callback function pthread signalling to prevent
  2282. deadlock with libusb's own event lock.
  2283. - Use a write lock when performing any USB control transfers to prevent
  2284. concurrent transfers.
  2285. - Free a libusb transfer after we have finished using it to avoid a dereference
  2286. in usb_control_transfer
  2287. - Do not perform bfi int patching for opencl1.2 or later.
  2288. - Although async transfers are meant to use heap memory, we never return before
  2289. the transfer function has completed so stack memory will suffice for control
  2290. transfers, fixing a memory leak in the process.
  2291. - klondike - correct/reverse min/max stats
  2292. - api incorrect message name
  2293. - klondike - use a link list queue rather than a circular buffer - and add
  2294. timing stats
  2295. - Use a timeout with usb handle events set to a nominal 200ms and wait for the
  2296. polling thread to shut down before deinitialising libusb.
  2297. - Use stack memory for hex used in stratum share submissions.
  2298. - Use stack memory in test_work_current, avoiding a malloc/free cycle each time.
  2299. - Provide a lower level __bin2hex function that does not allocate memory itself.
  2300. - Convert the bitfury driver to use the hash_driver_work version of hash_work.
  2301. - Add a hash_driver_work function to allow for drivers that wish to do their own
  2302. work queueing and management.
  2303. - Convert all usb control transfers to asynchronous communication with our own
  2304. timeout management as well.
  2305. - Klondike - increase circular read buffer size
  2306. - Klondike - extra zero value and range checking in temp conversion
  2307. - klondike - display MHz also
  2308. - Make pthread conditional timeouts handle all bulk usb transfer timeouts
  2309. performing libusb_cancel_transfer, disabling timeouts within libusb itself.
  2310. - Avoid calling get_statline_before on exit to avoid trying to use it on drivers
  2311. in an indeterminate state.
  2312. - Avoid calling get_statline on exit.
  2313. - Add a small amount to the usb timeout before cancelling to allow for a regular
  2314. usb polling interval to pass.
  2315. - Do not attempt to clear a usb halt before sending the cancel message since all
  2316. transfers should normally be cancelled before attempting to clear a halt
  2317. condition, and only change the return message to a timeout if it's consistent
  2318. with a cancellation.
  2319. - Retry up to USB_RETRY_MAX times to clear a halt condition before failing.
  2320. - Show the error number as well as the description in erroring bulk transfers.
  2321. - Drop logging level for failed to connect to stratum to verbose mode only since
  2322. we hit it regularly.
  2323. - We are always dependent on libusb handling events so use the blocking
  2324. libusb_handle_events in the polling thread and use a bool to know if we should
  2325. continue polling.
  2326. - Use fractional hashrate return values in bitfury_scanhash to minimise the
  2327. number of times we return 0 based on hashrate so far to further damp out
  2328. displayed hashrate.
  2329. - Check for presence of driver name in DRIVER_COUNT_FOUND to prevent strcmp on a
  2330. null pointer when a driver is not built in.
  2331. - CMR allow sending flash and clock commands
  2332. - Kill off threads that have failed using hash_sole_work instead of just
  2333. disabling them.
  2334. - Make the bf1 getinfo size a macro
  2335. - Failing to add_cgpu in bitfury should be a terminal failure.
  2336. - Check return values when attempting to open a BF1 device and set the msg size
  2337. as a macro.
  2338. - Display errors on failed usb read and write and consider sequential IO errors
  2339. a permanent failure.
  2340. - Use libusb's own error name function instead of hand coding the error names.
  2341. - Limit ms_tdiff to 1 hour as a sanity check.
  2342. - Enable the usb buffer in avalon driver.
  2343. - Check for async transfer variants of error messages.
  2344. - Remove unused variables.
  2345. - Try switching pools if for some reason we end up with only idle pools and have
  2346. ended up current_pool set to an idle one.
  2347. - Check a pool is stable for >5 mins before switching back to it.
  2348. - Minimise the time between dropping the read devlock and grabbing the write
  2349. devlock to avoid tons of logging spam in the interim.
  2350. - Check for libusb transfer stall error to be consistent with async IO errors
  2351. returned for a halt condition.
  2352. - Check for continuous IO errors on USB and consider the device inactive if more
  2353. than retry max.
  2354. - Make the devlock a cglock in usbutils and only grab the write lock for
  2355. fundamental changes allowing us to send and receive transfers concurrently
  2356. without lock contention.
  2357. - Prevent overflows in us_tdiff and ms_tdiff.
  2358. - Change second initialise message on bitfury verbose mode.
  2359. - Submitting an ntime offset nonce needs to be done on a copy of the work
  2360. instead of the original so abstract out shared components as much as possible,
  2361. minimising strdups in copy_work and make submit_work_async work take copied
  2362. work, cleaning up code in the process.
  2363. - Provide a way for drivers to submit work that it has internally rolled the
  2364. ntime value by returning the amount it has ntime rolled to be added.
  2365. - Typo in configure.ac
  2366. - Remove unmaintained broken ztex driver.
  2367. - Icarus - use a data structure for I/O rather than magic numbers
  2368. - delete old tracked ccan/opt/*.o files
  2369. - klondike correct cvtKlnToC() temperature calculation
  2370. - klondike - correct 1st reply debug based on define
  2371. - klondike - debug dump structured replies
  2372. - klondike - avoid division by zero if maxcount is unexpectedly zero
  2373. - klondike store and report errorcount and noise
  2374. - klondike - fix chipstats api stats buffer overrun with 16 chips
  2375. - klondike add new nonecount only once
  2376. - klondike - report mh/s based on nonces found + put old estimate into API stats
  2377. - klondike use a memcpy
  2378. - klondike fix bracket tabs indenting
  2379. - api.c missing Klondike from ASIC list
  2380. - Klondike update code to current git
  2381. - Add 2nd CMR to 01-cgminer.rules
  2382. - Add Klondike to 01-cgminer.rules
  2383. - Klondike to main directory
  2384. - Klondike consistent code spacing
  2385. - Klondike update driver code to current git
  2386. - update firmware for 16 chips, add dist files
  2387. - beta final 0.3.0 release
  2388. - updated firmware, IOC method
  2389. - prevent nonces when not state W
  2390. - added driver config option support
  2391. - fixes for 300 MHz, fix K1 parts list
  2392. - update driver, docs
  2393. - update firmware & utils
  2394. - updated cgminer driver for 3.3.1
  2395. - update firmware and driver, create new cgminer fork
  2396. - update klondike driver
  2397. - add cgminer driver file as-is
  2398. - Add API output displaying USB cancellations.
  2399. - Store statistics on how often we have to cancel async bulk transfers and add a
  2400. debug message whenever we do.
  2401. - Treat any unexpected timeouts waiting for async transfers as though there may
  2402. be a usb halt condition and attempt to clear the halt before cancelling the
  2403. tranfer.
  2404. - Remove zero packet flag on usb as it's unsupported outside linux and
  2405. unnecessary.
  2406. - Fake the libusb transfer timed out message if we force cancel it with our own
  2407. async functions.
  2408. - Use asynchronous transfers for all bulk transfers, allowing us to use our own
  2409. timers and cancelling transfers that take too long.
  2410. - Add libusb error warning message when significant error occurs.
  2411. - Icarus CMR2 detect FPGA setup
  2412. - Disable bitfury device thread on it disappearing.
  2413. Version 3.5.0 - 29th September 2013
  2414. - Add magic init sequence required on BF1 devices to get them mining on windows.
  2415. - usbinfo.devlock is only ever write locked so convert it to a mutex
  2416. - Icarus remove unneeded opt_debug tests due to applog being a macro
  2417. - Icarus - CMR shouldn't wait the full timeout due to handle sharing
  2418. - We should only yield once in cg_wunlock
  2419. - Provide a function to downgrade a cglock from a write lock to an intermediate
  2420. variant.
  2421. - Deuglify use of _PARSE_COMMANDS macro expansions.
  2422. - Deuglify use of usb parse commands macro in usbutils.
  2423. - Use the driver add commands macros in api.c to avoid individually listing
  2424. them.
  2425. - Separate out asic fpga and opencl drivers in the driver parse commands macro
  2426. for use individually as needed.
  2427. - Use macro expansion in usb_find_devices to avoid explicitly listing them all.
  2428. - Use macro expansion to iterate over all the drivers without explicitly writing
  2429. them out in usbutils.c
  2430. - Iterate over the bitfury offsets in order of decreasing likelihood.
  2431. - Reattach the kernel driver on linux on usb_uninit.
  2432. - Attach the kernel driver on failure to usb init on linux.
  2433. - libusb kernel driver operations are only available on linux.
  2434. - There is no need to get the external prototypes for drivers in cgminer.c any
  2435. more.
  2436. - Remove unnecessary gpu_threads initialisation.
  2437. - Put avalon last in the sequence of adding drivers to prevent it trying to
  2438. claim similar chip devices on startup.
  2439. - Use macro expansion to iterate over all device drivers without needing to
  2440. explicitly code in support in all places. Pass a hotplug bool to the detect()
  2441. function to prevent opencl trying to hogplug GPUs.
  2442. - Forward declare all device drivers in miner.h avoiding the need to export them
  2443. everywhere else.
  2444. - Add a noop function for driver detect when it's missing.
  2445. - Reuse the DRIVER_ macros to avoid having yet another definition for DRV_
  2446. - Use macro expansion to generate extern device_drv prototypes.
  2447. - Create a macro list of drivers to enable easier addition of further drivers.
  2448. - There is no point setting the BF1 preferred packet size to the maximum since
  2449. it will do so automatically.
  2450. - icarus ensure all cmr interfaces are initialised properly
  2451. - usbutils - fix USBDEBUG warnings
  2452. - Remove unnecessary steps in communicating with BF1 and just use USB interface
  2453. 1.
  2454. - usbutils - usb_bulk_transfer fix the buf/data fix
  2455. - usb_bulk_transfer - use the allocated buffer
  2456. - Set preferred packet sizes per interface on BF1.
  2457. - usbutils allow PrefPacketSize per endpoint
  2458. - Remove magic control sequences on open/close on BF1 and just flush the read
  2459. buffers.
  2460. - Check return codes in getinfo and reset and fail as needed in BF1.
  2461. - Check return code for bitfury_open and release resources properly on failed
  2462. initialisation.
  2463. - Abstract out flushing of interrupt reads in BF1 devices.
  2464. - Perform interrupt read after close message on BF1 as per serial close.
  2465. - Perform interrupt read flush as per serial open on BF1 devices.
  2466. - Add information for 2nd USB interface on BF1 devices and choose interface 1
  2467. for bulk transfers.
  2468. - usbutils - bulk transfer copy test fix
  2469. - usbutils - add USBDEBUG for usb_bulk_transfer
  2470. - Add more read_ii variants to usbutils.
  2471. - Name remainder of BFU usb commands used.
  2472. - Use submit_tested_work in bitfury driver to avoid unnecessarily re-testing the
  2473. work for validity.
  2474. - Abstract out work submission once it's been tested, to be used by drivers that
  2475. do their own internal validity testing.
  2476. - Store the hash2 array in struct work for further reuse.
  2477. - usbutils - which_intinfo not requried
  2478. - Use the test_nonce function within submit_nonce and store the uint32
  2479. corresponding to hash2 37 for further use.
  2480. - usbutils - interfaces must all be on one handle - ep implies the interface
  2481. - avalon stats use exact type
  2482. - Only set share diff if we've confirmed it's a share first.
  2483. - Update ASIC-README for bitfury devices.
  2484. - Use an array of offsets when checking nonces in bitfury_checkresults
  2485. - Limit the duration we wait for reads in BF1 based on time already elapsed to
  2486. account for other delays such as work restart messages or out of work.
  2487. - Minimise size of serial string we copy in BF1 stats to avoid overflow.
  2488. - Implement basic API stats for BF1 and increase array of results to check for
  2489. the rare straggling result.
  2490. - Space debug output for bf1 to separate from numerals.
  2491. - Abstract out the bitfury open close and reset functions and use them on
  2492. reinit.
  2493. - Rename BF1 devices BF1
  2494. - Check for work restart, breaking out early after usb reads in BF1.
  2495. - Do not lose the first sets of results from BF1.
  2496. - There is no point checking for results from the next round of work on BF1.
  2497. - Last result returned by BF1 is an end of results marker so ignore it.
  2498. - restart_wait should return 0 if thr_restart is true.
  2499. - Remove unused code by bitfury driver since current driver uses serialised
  2500. scanhash.
  2501. - Meter out return of estimated hashes in BF1 to smooth out visible hashrate.
  2502. - Optimise inner scanhash loop for bf1.
  2503. - Add yet another backup work for triple buffering of work in bf1 to account for
  2504. extra late results returned and don't check nonce offsets which appear to never
  2505. return.
  2506. - Name the work request and result usb commands for BF1
  2507. - Define a mandatory upper limit to waiting for reset and data on BF1 based on
  2508. full nonce duration.
  2509. - Decrease usb buffering to verbose logging.
  2510. - Add in first draft for a serialised work model sending/receiving data for BF1
  2511. devices.
  2512. - Add complete close sequence to bf1 as it happens on serial.
  2513. - Provide a bitfury identify function for bf1.
  2514. - Reliably extract BF1 information at startup and reset the device.
  2515. - Add commands for getting BF1 bitfury info
  2516. - Add magic BF1 bitfury open and close control sequences.
  2517. - Add BF1 detection code to bitfury driver.
  2518. - Create basic placeholders for bitfury driver code.
  2519. - Add bf1 device information to usbutils to enable device detection.
  2520. - Add basic defines for building for bitfury devices.
  2521. - Add redfury device to udev rules.
  2522. - avalon: display the FPGA controller version on API
  2523. - pool_active uninitialised_var rolltime
  2524. - Use macro expansion to only need to define usb enums and commands in one
  2525. place.
  2526. - usbutils saving incorrect overflow buffer
  2527. - ignore libusb.la and *.lo on linux
  2528. - icarus support CMR with no extensions
  2529. - usbtils - interfaces dont work yet in libusb windows so disable for that only
  2530. - Provide a --disable-libcurl config option to build support for stratum mining
  2531. only.
  2532. - Fix the api-example.c compile under Linux
  2533. - usbutils - only release the device once - for the first intinfo
  2534. - usbutils set_interface is no longer valid
  2535. - ubsutils interfaces much each have their own handle
  2536. - usbutils kernel_detach should use the interface number
  2537. - usbutils - allow the driver to change which_intinfo
  2538. - Reset quotas on load balance for all pools at the same time to avoid running
  2539. out during selection and unintentionally dropping to fallback.
  2540. - Break out of select pool from a common point for appropriate debug messages
  2541. and to avoid further tests.
  2542. - usbutils correct/reverse CMR product numbers
  2543. - usbutils specifically track handles and interfaces
  2544. - change drivers to use usb_interface() - required for multi interface change
  2545. - usbutils - allow a device to use multiple interfaces (and better var names)
  2546. - Cast -1 to (char) to cope with different default char types on ARM.
  2547. Version 3.4.3 - 13th September 2013
  2548. - Put corefoundation and iokit separate in ldflags for darwin.
  2549. - Add rules for libusb Makefile.am building on osx
  2550. - Add flags for building libusb statically on osx.
  2551. - Find the greatest common denominator in quotas and use the smallest number of
  2552. consecutive work items per pool in quota load balance mode to smooth hashrate
  2553. across pools with large quotas. Give excess quota to priority pool 0 instead of
  2554. pool 0.
  2555. - Avoid dynamically adding stack memory for nonce2 in the stratum send thread
  2556. and check the pool's nonce2_len will not cause an overflow.
  2557. - Add subdir-objects to automake options.
  2558. - Use inet_addr instead of inet_network to fix windows build.
  2559. - Remove unused pbase variable.
  2560. - Add support for socks4/4a proxies with stratum, and drop back to socks4
  2561. support via the global --socks-proxy command to not break previous
  2562. configurations.
  2563. - Fix warning on mingw build.
  2564. - Only show long-poll message in pool summary if it's not using stratum.
  2565. - Increase the time for the waiting for work message to be given to be greater
  2566. than that required for a pool swap in the scheduler which is set to 5s.
  2567. - Change message in status when using a balanced pool strategy to notify if
  2568. there's a stratum pool as well.
  2569. - Use the --failover-only flag to have special meaning in combination with
  2570. load-balance mode to distribute any unused quota back to pool 0 to maintain
  2571. ratios amongst other pools.
  2572. - Display quota and allow it to be modified via the pool menu.
  2573. - Add API commands and modify output to support pool quota displaying and
  2574. changing.
  2575. - Change message in status when using a balanced pool strategy to notify if
  2576. there's a stratum pool as well.
  2577. - Add quota support to configuration files.
  2578. - Rotate pools on all failures to set a pool in select_pool.
  2579. - Use quotas for load-balance pool strategy.
  2580. - Provide a mechanism for setting a pool quota to be used by load-balance.
  2581. - Use the --socks-proxy option with stratum, changing it to defaulting to socks5
  2582. and give appropriate message should it fail to connect.
  2583. - Cope with trailing slashes in stratum urls.
  2584. - Add more debugging messages when negotiating with proxies for stratum.
  2585. - Test specifically for socks5h in socks support for stratum.
  2586. - Add support for socks5 proxy with stratum
  2587. - Provide support for negotiating a stratum connection via http proxies.
  2588. - Connect to the proxy URL and port if specified for stratum sockets instead of
  2589. the pool directly.
  2590. - Extract any proxy url and port to be used by sockaddr if possible using
  2591. extract_sockaddr.
  2592. - Make extract_sockaddr set variables passed to it rather than pool struct
  2593. members.
  2594. - miner.php sort the mcast rigs so they are always in the same relative order
  2595. - miner.php allow sending the muticast message multiple times
  2596. - miner.php mcast ignore duplicate replies
  2597. Version 3.4.2 - 3rd September 2013
  2598. - take_queued_work_bymidstate should use a write lock.
  2599. - miner.php coding warning
  2600. - miner.php disable 'gen' by default
  2601. - miner.php allow formula generation of new fields
  2602. - miner.php add doctype
  2603. - miner.php remove incorrect echo
  2604. - miner.php optional error if not enough mcast rigs are found
  2605. Version 3.4.1 - 31st August 2013
  2606. - API mcast add a description option with miner.php
  2607. - Always use a maxpacketsize buffer in usb_bulk_transfer
  2608. - bflsc ensure getinfo cannot overflow it's storage buffer
  2609. - Don't decref json values in stratum parsing due to memory corruption.
  2610. - Use 64 bytes for all libusb control transfers.
  2611. - Skip dissecting opt->names in parse_config if it doesn't exist.
  2612. - Use an internal buffer in _usb_transfer_read in case the read is larger than
  2613. the buffer passed to it.
  2614. - ICA optional limit timing with short=N or long=N
  2615. - Revert to old custom tolines function since strtok_r is not portable.
  2616. - bflsc remove unused commented out code
  2617. - logging - code mistake
  2618. - logging - applogsiz() for large messages
  2619. - Provide base structures for getaddrinfo.
  2620. - Include string.h in bflsc driver.
  2621. - Get rid of linear removal of spaces in bflsc text parsing and use strstr
  2622. throughout instead.
  2623. - Use reentrant strtok in tolines() function in bflsc to avoid racing on
  2624. contextless calls.
  2625. - Show how small a too small result in bflsc is.
  2626. - Duplicate the buffer in process_results in bflsc since strtok modifies it
  2627. making debugging output limited to one line.
  2628. - Only process nonces in bflsc if the breakdown function succeeds.
  2629. - Ignore zero count messages in bflsc instead of trying to parse them.
  2630. - Return ok in tolines when it doesn't match inprocess message for bflsc.
  2631. - Remove inprocess line instead of deleting all following responses in bflsc.
  2632. - Change ok testing logic in breakdown() in bflsc and return if not ok at any
  2633. stage.
  2634. - Check the return value of tolines in bflsc driver.
  2635. - Use strtok to parse lines in bflsc driver.
  2636. - Add libusb-1.0 m4 directory and gitignore file.
  2637. - Properly convert from ranlib to lt_init in configure.ac
  2638. - Make autoconf always build for libusb.
  2639. - More autoconf fixes.
  2640. - Unconditionally build jansson statically from the cgminer source tree.
  2641. - Only test for all usb devices once in configure.ac
  2642. - Fix various libusb warnings and possible bugs on linux build.
  2643. - Add make clean and maintainer-clean to autogen
  2644. - Remove examples from libusb Makefile and generated autoconf files.
  2645. - Fix libusb subdirectory builds.
  2646. - Remove cached files from libusb autoconf on running autogen.sh
  2647. - Remove unused HAVE_LISBUSB macro and use USE_USBUTILS everywhere.
  2648. - Use direct auto* files to avoid failure of autoreconf
  2649. - Remove unused and maintainer cleaned files
  2650. - Show RT_LIBS in ./configure output.
  2651. - First import of libusb-1.0
  2652. - bflsc xlinkstr use snprintf
  2653. - Fix win32 build.
  2654. - Use take_queued_work_bymidstate in the bflsc driver to avoid the rare chance
  2655. repeated results come back from the same work item.
  2656. - Provide a funcion that looks up queued work by midstate and then removes it
  2657. from the device hash database.
  2658. - Fix no -rt library on darwin.
  2659. - Update included jansson to v2.4
  2660. - Fix OSX build.
  2661. - Provide an osx fix for cgtimers and a fallback to timevals for all other
  2662. platforms !linux !win32 !osx.
  2663. - Move two more timer functions out of define macros to enable them to be used
  2664. by future osx code.
  2665. - cgtimer_sub is now the same since cgtimer_t should be the same on all
  2666. platforms.
  2667. - miner.php fix missing global
  2668. - Only count submitted nonces as diff1shares if they're valid.
  2669. - Substantially raise the maximum avalon frequency for water-cooled, over-volted
  2670. designs.
  2671. - Compile MCast.java with an old java
  2672. - API Multicast sample MCast.java+MCast.class
  2673. - BTB show C/MHz/mV for device
  2674. - api.c remove unused reply string
  2675. - api.c fix mcast debug message bug
  2676. - miner.php implement API Multicast handling to automatically find your local
  2677. net miners
  2678. - API mcast only reply to remote IP's that are allowed access
  2679. - Initial API Multicast response v0.1 to find cgminer APIs
  2680. - Use timespecs on windows as cgtimer_t to capitalise on the higher resolution
  2681. clock changes.
  2682. - Abstract out the conversion of system time to an lldiv_t in decimicroseconds.
  2683. - Use our own gettimeofday implementation on windows for it to be consistent
  2684. across ming builds and higher resolution.
  2685. Version 3.4.0 - 21st August 2013
  2686. - Use stack data for HW error% in avalon stats.
  2687. - Add avalon HW error% to stats and only show BTB variables if avalon is a BTB.
  2688. - Check for cnx_needed on each loop through wait_lp_current.
  2689. - Return positive for cnx_needed when no_work is true.
  2690. - Stratum is used more often so test for it first.
  2691. - Reorder support names alphabetically.
  2692. - Only display the no pool work message once if there are multiple waiters in
  2693. hash_pop
  2694. - Provide a message and set a bool when no work is available from any pools and
  2695. when it resumes again.
  2696. - We don't want to continue into the hash_pop function if the getq is frozen.
  2697. - Only report threads in and out in queued work devices across a get work since
  2698. the rest happens asynchronously and the get work is what the device might be
  2699. waiting on.
  2700. - Thread reportin and out can be static non inline.
  2701. - usbutils cps sleep_estimate is not an underestimate
  2702. - usbutils add cps stats estimates
  2703. - Provide cgtimer_sub helper functions.
  2704. - Provide cgtimer_to_ms helper functions.
  2705. - Rename cgsleep_prepare_r as cgtimer_time to get time in cgtimer_t format and
  2706. call cgsleep_prepare_r as a macro for cgtimer_time
  2707. - Use the reentrant cgsleep functions for usecps in usbutils.
  2708. - TimeBeginPeriod and TimeEndPeriod do not add significant overhead when run the
  2709. entire time for cgminer so avoid trying to maintain balanced numbers of them for
  2710. specific time calls to simplify code.
  2711. - Replace all references to the old n*sleep functions with the equivalent
  2712. cgsleep_*s replacements.
  2713. - timeGetTime uses huge resources on windows so revert to using timevals for its
  2714. implementation of cgtimer_t
  2715. - Quotient/remainder error in ms division.
  2716. - Only grab a queued work item if we successfully grab the lock to submit work
  2717. in bflsc_send_work
  2718. - BTB get version from Firmware
  2719. - Carve out the unused portions of sha2 implementation.
  2720. - Import Aaron D. Gifford's fast sha256 implementation.
  2721. - Increase the que_low watermarks on BFLSC for they are too low to keep the
  2722. device busy on scanwork loops.
  2723. - Provide cgtimer_to_timeval helper functions.
  2724. - Provide a timeval_to_cgtime helper function to reuse values.
  2725. - Check for thr->work_restart in restart_wait.
  2726. - We should be using que_low to decrease scan sleep time in bflsc.
  2727. - Prepare sleep time on bflsc if no dev needs work yet to avoid busy waiting.
  2728. - Simplify cgsleep code for windows by using a typedef for cgtimer_t that
  2729. resolves to clock resolution, using that internally.
  2730. - On windows use the higher accuracy timegettime function to really get 1ms
  2731. clock and timer accuracy.
  2732. - Use the cgsleep reentrant function to sleep for bflsc between read results to
  2733. account for time taken to perform reads.
  2734. - Use 100ms delay between checking for results on all bflsc devices as the
  2735. buffering of results mean checking more frequently just wastes CPU and causes
  2736. more lock contention for only marginally better latencies.
  2737. - Fix missed endtimeperiod in overrun timer on windows.
  2738. - Make cgsleep_us_r take an int64_t for us.
  2739. - Make the cgsleep functions build on windows.
  2740. - Use the cgsleep reentrant function in avalon_send_task.
  2741. - Use the reentrant cgsleep functions within the avalon_send_tasks function.
  2742. - Set high resolution timing on windows within the cgsleep functions.
  2743. - Use the reentrant cgsleep function to time sleeps on reading from avalon.
  2744. - Provide reentrant versions of cgsleep functions to allow start time to be set
  2745. separately from the beginning of the actual sleep, allowing scheduling delays to
  2746. be counted in the sleep.
  2747. - Make the nmsleep and nusleep functions use the new cgsleep functions
  2748. internally till functions are migrated to the new cgsleep API.
  2749. - Add a ms_to_timespec helper function, and create a cgsleep_ms function that
  2750. uses absolute timers with clock_nanosleep to avoid overruns.
  2751. - Add rt lib linkage to enable use of clock_nanosleep functions with older
  2752. glibc.
  2753. - Add necessary time header include to avalon driver.
  2754. - Do a sleep of the full duration it would take to do all the work using
  2755. clock_nanosleep in avalon_send_tasks to avoid sleep overruns before polling to
  2756. see if it's ready.
  2757. - Add a timeraddspec helper function.
  2758. - Provide a us_to_timespec helper function.
  2759. - Use the us_to_timeval helper function in the avalon driver.
  2760. - Provide a us_to_timeval helper function.
  2761. - Use timeval_to_spec helper in avalon driver.
  2762. - Add helper functions to convert timespec to timeval and vice versa.
  2763. - simplifying buffer full check
  2764. - forking bitburner write thread function
  2765. - making sure original Avalon is unaffected by BitBurner changes
  2766. - changes to queueing strategy for BitBurner boards
  2767. - Do not poll in avalon_get_results without sleeping if we have finished parsing
  2768. a full result.
  2769. - Add c to ambient temperature display for avalon driver.
  2770. - BTB allow up to 1400mV as per firmware limits
  2771. - avalon for timeout allow d='calculate it' and fix uninitialised
  2772. - Use cloned work when finding avalon results since another thread can discard
  2773. the work item while it's in use.
  2774. - Provide a variant of find_work_bymidstate that returns a clone of the found
  2775. work.
  2776. Version 3.3.4 - 14th August 2013
  2777. - API/miner.php add some % fields
  2778. - Nonce2 stratum submission is not working with nonce2 lengths >4, revert the
  2779. buggy __bin2hex function and use bin2hex.
  2780. - The write thread in avalon is only ever actually woken up by timeout so remove
  2781. the write semaphore and use a simple sleep poll.
  2782. - Fix warning.
  2783. - Interrupting reads on the avalon to start writes loses data so remove the
  2784. cgsem_post in the read code.
  2785. - Add room for the null byte at the end of the nonce2 string on stratum share
  2786. submission and zero the allocated ram.
  2787. Version 3.3.3 - 13th August 2013
  2788. - Only perform the bin2hex on nonce2 data if it's required for stratum
  2789. submission, thereby removing the last conversion of that type from stratum work
  2790. generation.
  2791. - Create a work data template when receiving stratum notification, allowing a
  2792. simple memcpy of the merkle root avoiding more hex2bin conversions on each work
  2793. generation.
  2794. - Export the workpadding char in miner.h
  2795. - Avoid a potential overflow should a pool specify a large nonce2 length with
  2796. stratum.
  2797. - Avoid one more hex2bin in gen stratum work.
  2798. - Rename work gbt_coinbase to coinbase to be in line with pool variable name.
  2799. - Perform merkle bin hex2bin on stratum notify to avoid doing it on each work
  2800. generation.
  2801. - Reuse just the one pool coinbase variable in stratum, avoiding more string
  2802. functions and storage in gen_stratum_work on each work generation.
  2803. - Rename pool gbt_coinbase variable to coinbase to combine it with the stratum
  2804. coinbase data.
  2805. - Use a nonce2 offset variable for both gbt and stratum to consolidate
  2806. requirements on work generation.
  2807. - Merge pull request #474 from kanoi/master
  2808. - util.c update quit call for new functions
  2809. - use correct define for OSX in util.c
  2810. - miner.h inline semaphores increase information on failure
  2811. - util.c expand quit to show file/func/line
  2812. - Merge remote-tracking branch 'conman/master'
  2813. - Cache as much of the gbt coinbase as possible to avoid doing unnecessary
  2814. hex2bin conversion on every work generation with gbt.
  2815. - We should be using a cg_wlock initially in generating stratum and gbt work
  2816. before downgrading the lock.
  2817. - Add the ability to downgrade a write variant of the cglocks.
  2818. - Fix --scrypt being required before scrypt intensities on command line or not
  2819. working at all via config files.
  2820. - Cache the hex2bin of pool nonce1 in stratum, avoiding hex2bin on each work
  2821. generation.
  2822. - Cache the binary generation of coinbase1 and 2 on stratum, avoiding a hex2bin
  2823. of coinbase1 and 2 on each work generation.
  2824. - cgsem - increase information on failure
  2825. - avalon init write_sem before use
  2826. - 3.3.2 - 9th August 2013
  2827. - Fix uninit variable warnings.
  2828. - usbutils - force check every combination
  2829. - Fix warning.
  2830. - Recreate curses windows on windows when a device is hotplugged to allow window
  2831. resizing without crashing.
  2832. - Update copyright notice.
  2833. - Limit intensity range according to whether scrypt is in use or not.
  2834. - Do not allow benchmark mode to be used with scrypt.
  2835. - Add a --bflsc-overheat command which allows you to set the throttling
  2836. temperature for BFLSC devices or disable it.
  2837. - Move bflsc defines to a header file.
  2838. - avalon allow frequency to be set via the API
  2839. - BTB voltage management via the API - and set default on startup
  2840. - Avalon BTB allow partial work to be transferred
  2841. - avalon_cts use correct buffer
  2842. - miner.php format Best Share
  2843. - remove unnecessary memcpy
  2844. - using more concise description
  2845. - using usb_ident
  2846. - forgot a return
  2847. - changes to Avalon driver for BitBurner boards
  2848. - Revert "Sleep after sending icarus work to emulate working at 115200 baud."
  2849. - api correct timeout stat display
  2850. - usb timeouts - min/max also
  2851. - log USB timeouts in API stats
  2852. - usbutils report failed timeouts
  2853. - usbutils ensure stats macros are using the macro arguments
  2854. - Check for negative wait time in socket_full.
  2855. - Fix extra argument passed to statline before.
  2856. - Adjust socket wait timeout in recv_line according to how long we've already
  2857. waited to avoid a 60 second wait dropping to 1 second due to a blocked socket.
  2858. - usbutils use a heap buffer for bulk read rather than stack
  2859. - usbutils only one bulk transfer call per stat
  2860. - set device_drv function noops when first add_cgpu
  2861. - usbutils - in init only change the config if needed
  2862. - bflsc nonce per work item stats
  2863. - bflsc increase flush count to handle parallel work
  2864. - force type checking on curses
  2865. - logging - size check sprintf
  2866. - usbutils - size check all sprintf
  2867. - cgminer - size check all sprintf
  2868. - size check get_datestamp/get_timestamp and remove unused cgpu->init
  2869. - make all statline overflow safe
  2870. - WU only needs +2 width
  2871. - Check for a timeout in avalon_scanhash and post to the write sem if we receive
  2872. one.
  2873. - Decay result count in avalon more slowly to not falsely detect idle periods as
  2874. low result return rates.
  2875. - Count the number of miners idled in avalon to account more accurately for when
  2876. its result return rate is too low.
  2877. - Fix potential dereference when starting avalon with all new work.
  2878. - Convert the decay_time function into one that truly creates an exponentially
  2879. decaying average over opt_log_interval.
  2880. - Only throttle avalon clockspeed in avalon_auto in non optimal temperature
  2881. settings if the fanspeed has reached maximum.
  2882. - Reinstate more aggressive <2% HW error target for avalon-auto
  2883. - Set avalon fan min and fan max to PWM values instead of percentage.
  2884. - Provide an --avalon-freq command line to give a valid range of frequencies for
  2885. avalon in auto mode.
  2886. - Set the avalon idle frequency to lowest if avalon auto is enabled and we have
  2887. an overheat condition.
  2888. - Decrease avalon frequency in auto mode if we are unable to maintain the
  2889. temperature in the optimal range.
  2890. - Don't count invalid nonces as hashrate for bflsc.
  2891. - Use a more conservative upper limit of 1% for hardware errors with avalon auto
  2892. frequency.
  2893. - Allow the avalon fanspeed range to be passed as parameter on the command line,
  2894. default to 20-100%
  2895. - Just display A: and R: for difficulty accepted and rejected to preserve screen
  2896. real estate and decrease decimal places for WU.
  2897. - correct device DR: and remove global U:
  2898. - Update all screen A/R to instead use DA/DR and device U to WU
  2899. - miner.php add ASC fields
  2900. - GPU fan rpm display 9999 when it overflows
  2901. - bflsc get volts stats needs its own GETVOLTS
  2902. - Support all avalon frequencies on the command line.
  2903. - Move to slightly more relaxed timeouts for avalon.
  2904. - MMQ turn on cps delays
  2905. - bflsc x-link header different to documentation
  2906. - Reset the other auto counters in avalon when idling a device.
  2907. - usbutils/icarus include more locking to usbdev access
  2908. - Icarus turn on cps delays by default
  2909. - usbutils cps correct time measurement
  2910. Version 3.3.1 - 25th June 2013
  2911. - Add an avalon-auto option which enables dynamic overclocking based on hardware
  2912. error rate for maximum effective hashrate.
  2913. - Add an --avalon-cutoff feature which puts the avalon idle should it reach this
  2914. temperature, defaulting to 60, re-enabling it when it gets to target
  2915. temperature.
  2916. - Change default avalon target temperature to 50 degrees.
  2917. - usbutils - incorrect test for * in bus:dev
  2918. - Redo +1 fix in bflsc.
  2919. Version 3.3.0 - 24th June 2013
  2920. - Add an --avalon-temp option to allow a user specified target temperature.
  2921. - Demote no matching work message to verbose logging only on avalon.
  2922. - Make the fan control on the avalon a simple PID controller with a target
  2923. temperature of 45.
  2924. - Demote bflsc hw error messages to verbose logging only.
  2925. - bflsc - handle xlink timeouts by having generic IO functions
  2926. - Demote the invalid nonce warning to log info.
  2927. - Ignore iManufacturer for BFLSC devices since the device name will still match
  2928. and some unbinned chips are missing it.
  2929. - sc_count shouldn't be +1 in bflsc.
  2930. - Use the info timeout for read_nl in getidentify bflsc.
  2931. - Add a usb_read_nl_timeout macro.
  2932. - bflsc try getinfo twice
  2933. - set MSG_ASCUSBNODEV always defined
  2934. - Hard code the preferred packet size for AMU, BLT and ICA.
  2935. - API V1.26 update ASIC support
  2936. - Icarus enable the read buffer for the detect nonce
  2937. - Support new overclocking speeds for avalon: 325, 350 and 375
  2938. - undo icarus show errno, put it as debug in ubsutils
  2939. - icarus add errno to rerr and werr
  2940. - Sleep after sending icarus work to emulate working at 115200 baud.
  2941. - Use the nusleep function for sleeping after sending work in avalon.
  2942. - Show an integer only for diff if it is one.
  2943. - Set the avalon preferred packet size to 512.
  2944. - Reinstate the maxPacketSize determined by the end descriptor but allow the
  2945. driver to override it.
  2946. - Only update hashmeter if we have done hashes or haven't updated longer than
  2947. the log interval, fixing a us/ms error.
  2948. - Use only one cgsem in avalon signalling when the write thread should commit
  2949. work by reading the status bytes off during an avalon_read, minimising the
  2950. number of usb calls and resetting from only one place.
  2951. - Change avalon no valid work message to no matching work to match API
  2952. terminology.
  2953. - Use low latency usb transfers on the avalon, sleeping up to half a buffer's
  2954. worth only if no data is returning to increase hashrate, abolish lost work and
  2955. decrease CPU.
  2956. - Minimise the sleep times in avalon read to avoid result loss.
  2957. - Use a half nonce range before cycling through avalon's scanwork to ensure it
  2958. gets a chance to fill work if time is tight for the write thread to signal a
  2959. wakeup.
  2960. - Temporarily limit usb transfer sizes to 512 till we provide a way for each
  2961. driver to choose the upper limit.
  2962. - Increase watchdog sick time to longer than it takes for a pool to be detected
  2963. dead.
  2964. - Limit USB transfers to the max size reported by the descriptors.
  2965. - Increase the BFLSC timeout to allow the maximum number of results to be
  2966. returned for BAS in time.
  2967. - Decrease BAL and BAS latency to be just larger than one result read.
  2968. - disable curses device resize that crashes on windows
  2969. - BFLSC latest firmware has its own thermal cutoff set to 90, so use the same
  2970. value in case we have an old firmware that isn't throttling by itself.
  2971. - Drop watermark low limits for bflsc.
  2972. - Set the fanspeed on bflsc to max if we don't know the temperature.
  2973. - Use a low watermark for queueing mandatory work on bflsc instead of zero.
  2974. - Only mandatorily grab the bflsc mutex on submitting work when the queue is
  2975. empty.
  2976. - Adjust bflsc v2 watermarks.
  2977. - Only increase sleep time on bflsc if the queue isn't emptying at all over the
  2978. sleep duration.
  2979. - Fix warning.
  2980. - bflsc yet more API stats
  2981. - bflsc add some more API stats
  2982. - bflsc correct firmware matching
  2983. - bflsc correct comment
  2984. - Fixed Commands with No params
  2985. - bflsc driver support for v2 firmware
  2986. - Odd Issues
  2987. - Fixed Python Example
  2988. - Added Python Api Example
  2989. - Added Python Api Example
  2990. - Multiplier fail for microseconds vs milliseconds when updating hashmeter in
  2991. hash_queued_work.
  2992. - Only make threads report in/out across the actual driver code and update their
  2993. status on reporting out as well as in.
  2994. - usbutils initialise close key/sem
  2995. - usbutils cleanup linux semaphores on release
  2996. - Difficulty should be unconditionally byteswapped, not swapped to big endian.
  2997. - We should be setting cancelstate, not canceltype when disabling it for usb
  2998. locking.
  2999. - Pthread cancel state should be set to disable on usb DEVLOCK.
  3000. - Fanauto on bflsc is Z9X according to the source code, not 5 as per the draft
  3001. protocol document.
  3002. Version 3.2.2 - 16th June 2013
  3003. - Record and report USB pipe errors via API stats
  3004. - Suspend stratum connections when we know they've failed and don't try to recv
  3005. data from them once the socket no longer exists.
  3006. - Pipe error is quite common on usb3 so drop logging to verbose level only.
  3007. - ocl.c fix applog warnings on windows
  3008. - applog/quit fix GPU errors created
  3009. - usbutils - DEVLOCK other usbdev access
  3010. - applog usb device list can be > LOGBUFSIZ
  3011. - fix windows log warnings
  3012. - logging remove extra added <LF>
  3013. - remove varargs from logging/quit/in general as much as possible
  3014. - Don't yield when downgrading a cg ilock.
  3015. - Don't yield on grabbing the read lock variant of cglocks.
  3016. - Off by one error in device count for display.
  3017. - Don't display devices beyond the most_devices count in the curses status.
  3018. - Only display as many device rows as the maximum live existed at any time.
  3019. - usb lock out use cg locks
  3020. - usb lock out transfers during open/close
  3021. - Add error message to libusb pipe error
  3022. - Differentiate libusb control transfer pipe errors from transfer errors since
  3023. they're not fatal.
  3024. - Create a usb_bulk_transfer wrapper for libusb_bulk_transfer to cope with pipe
  3025. errors.
  3026. - Only show efficiency in pool information for pools that don't support local
  3027. work generation.
  3028. - Create a pool_localgen bool function for testing when a pool can generate work
  3029. locally.
  3030. - ignore file that is generated on Macs
  3031. - compile unix code on Mac OS X fixes not finding the config file in $HOME
  3032. - Use mining start time for device MH/U calculations
  3033. - Decrease the sleep duration before reading in avalon to not let the read
  3034. buffer overflow.
  3035. - Failure to read and write on pseudo semaphores on apple happens routinely on
  3036. shut down so should not be a quit error, just a warning.
  3037. - Unlock usb dev lock in the same place in usbutils.
  3038. - Sleep if the avalon buffer is empty and we've requested a read to allow the
  3039. write thread to take precedence.
  3040. - Yield after releasing a lock in case we are on a device with limited CPU
  3041. resources.
  3042. - Add the cgpu_info structure before avalon reset.
  3043. - Tidy up DEVLOCK/UNLOCK to have consistent use of the pstate variable without
  3044. needing brace level match.
  3045. - Icarus driver elaspsed timeout shouldn't be just USB I/O
  3046. - usbutils avoid leaving devlock locked when thread cancelled
  3047. - MMQ fix nodev failure caused by changes
  3048. - ubsutils lock all access to nodev and cgusb
  3049. - USB make device_path handled by usbutils
  3050. - tidy up free in device detect functions
  3051. - USB control creation and free of cgpu
  3052. - Add FAQ regarding Work Utility.
  3053. - Throttling the BFLSC at 80 seems to prevent generating garbled responses of
  3054. higher temps.
  3055. - Return after failed bin2hex conversion in bflsc.
  3056. - Demote failed hex2bin result to LOG_INFO and check return result in
  3057. driver-bflsc to avoid doing find_work_by_midstate.
  3058. - Set BFLSC fan speed coarsely to keep it under 60 or auto as per specs saying
  3059. it tries to stay below 60.
  3060. - Limit usbutils LATENCY_STD to 32ms to keep transfers under 512 bytes.
  3061. - Move macro definition to bflsc driver
  3062. - Use a longer timeout for retrieving bflsc details.
  3063. - Add a usb_read_ok_timeout wrapper to cope with slow init'ing devices.
  3064. - cgsem_post after creating the thread info
  3065. - Fix build.
  3066. - Use cgsem structures instead of the flaky pings in the work queue to start
  3067. mining threads and remove the unused thr_info_freeze function.
  3068. Version 3.2.1 - 7th June 2013
  3069. - Shorten the avalon statline to fit in the curses interface and show the lowest
  3070. speed fan cooling the asic devices.
  3071. - Set usbdev in usbutils after checking for nodev to avoid trying to access a
  3072. dereferenced value.
  3073. - AMU usbstatus correct name from enable UART
  3074. - Icarus AMU enable the UART
  3075. - Only libusb close if libusb release succeeds.
  3076. - Failed reads and writes on cgsem_post and cgsem_wait should be extremely rare.
  3077. - Implement cgminer specific cgsem semaphores to imitate unnamed semaphore
  3078. behaviour on osx which does not support them.
  3079. - Set cgusb->buffer to NULL when doing usb_buffer_disable.
  3080. - Temporarily fix apple not having semtimedop by ignoring the timeout value.
  3081. - BFLSC enable buffered USB reading
  3082. - Icarus use buffered USB reading
  3083. - bflsc & icarus use usb_ftdi_set_latency
  3084. - usb_ftdi_set_latency LOG_ERRs if called incorrectly
  3085. - add usb_ftdi_set_latency
  3086. - usbutils optional read buffering
  3087. - Set the avalon read transfer latency to avoid sleeping when no data is
  3088. returned after very short latency settings.
  3089. - correct bflsc BFLSC_BUFSIZ max calculation
  3090. - Fix build for !curses
  3091. - restore max code - since timeout is unsigned
  3092. - compile warning - remove unused max
  3093. - usb set FTDI latency higher to minimise status bytes
  3094. - Check for zero timeout on _usb_write.
  3095. - Check for zero timeout in usb read.
  3096. - Define a minimum polling time based on frequency of mandatory updates of ftdi
  3097. responses at 40ms.
  3098. - Sleep right up to the timeout instead of the first half if we find ourselves
  3099. polling in _usb_read
  3100. - Enforce half timeout sized sleeps in usb_read if we find the device is not
  3101. respecting libusb timeouts to avoid polling frequently.
  3102. - Add more ASIC documentation.
  3103. - Update README
  3104. - Remove start device limitation on log window size to allow it to get larger
  3105. with hotplugged devices.
  3106. - Switch logsize after hotplugging a device.
  3107. - Change switch_compact function name to switch_logsize to be used for other
  3108. changes.
  3109. - Only adjust cursor positions with curses locked.
  3110. - devs display - fix GPU duplicate bug
  3111. - Do not hotplug enable a device if devices have been specified and the hotplug
  3112. device falls outside this range.
  3113. - Change the --device parameter parsing and configuration to use ranges and
  3114. comma separated values.
  3115. - basic copyright statement in API.java
  3116. - devs display - show ZOMBIEs after all others
  3117. - Modify scrypt kernel message.
  3118. - Check for pool_disabled in wait_lp_current
  3119. - usbutils semun use proper def for linux which fixes OSX also
  3120. - Check for pool enabled in cnx_needed.
  3121. - Icarus add delays during intialisation
  3122. - Update documentation.
  3123. - Update copyrights of modified files.
  3124. Version 3.2.0 - 31st May 2013
  3125. - Add FAQ about windows USB keyboards and hotplug interactions.
  3126. - Fix mingw build warnings in icarus driver.
  3127. - Make usb_ftdi_cts use the _usb_transfer_read function.
  3128. - Update ASIC-README with avalon info regarding default behaviour.
  3129. - Break out of idling loop in avalon_idle if the buffer is full.
  3130. - Provide some defaults for avalon if none are specified and do not try to claim
  3131. the device if it fails to reset with them and no options are specified.
  3132. - usbutils automatically track IO errors
  3133. - usbutils allow a short wait for resources to be released
  3134. - correct semaphore timeout comment
  3135. - Set the fanspeed to the nominal chosen for GPUs.
  3136. - Inverted sem_init logic.
  3137. - Document avalon options in ASIC-README
  3138. - Do avalon driver detection last as it will try to claim any similar device and
  3139. they are not reliably detected.
  3140. - Clamp initial GPU fanspeed to within user specified range.
  3141. - Use a counting semaphore to signal the usb resource thread that it has work to
  3142. do.
  3143. - Avalon fan factor is already multiplied into the info values.
  3144. - Get rid of zeros which corrupt display.
  3145. - Logic fail on minimum fanspeed reporting.
  3146. - Provide a workaround for fan0 sensor not being used on avalon and pad fan RPM
  3147. with zeros.
  3148. - Add ambient temp and lowest fan RPM information to avalon statline.
  3149. - Display max temperature and fanspeed data for avalon.
  3150. - Set devices to disabled after they exit the hashing loops to prevent the
  3151. watchdog thread from trying to act on them.
  3152. - Add avalon driver to hotplug.
  3153. - Shut down the avalon mining thread if the device disappears.
  3154. - Check for no usb device in usb_ftdi_cts
  3155. - Check for valid usbdev in _usb_read in case the device has been unplugged.
  3156. - Scanhash functions perform driver shutdown so don't repeat it.
  3157. - Change the opencl shutdown sequence.
  3158. - Send the shutdown message to threads and do the thread shutdown functions
  3159. before more forcefully sending pthread_cancel to threads.
  3160. - Use the cgpu_info shutdown to determine when to stop the avalon read and write
  3161. threads.
  3162. - Use semaphores to signal a reset to pause the read thread while the write
  3163. thread does the actual reset, making all writes come from the same place.
  3164. - Remove now unneeded fgpautils.h include from avalon.
  3165. - usb_transfer_read should also not play with the endianness.
  3166. - Use the USB wrappers for avalon, telling usbutils that we want the raw data.
  3167. - Use separate ep for avalon tasks vs avalon reset and do not loop in write
  3168. indefinitely.
  3169. - Remove unneeded function and checks in avalon write code.
  3170. - CMR handle baud options
  3171. - work_restart is reset within the queued hash work loop.
  3172. - Fix avalon shutdown sequence.
  3173. - Execute driver shutdown sequence during kill_work.
  3174. - Use nusleep in avalon_get_results in place of nmsleep.
  3175. - Provide an nusleep equivalent function to nmsleep.
  3176. - usb/ica add more (incomplete) CMR settings
  3177. - Give a buffer of perceived results in avalon during idle periods to allow for
  3178. results once it becomes active again.
  3179. - libusb_control_transfer are meant to be endian specific, but host endianness
  3180. so no conversion is needed.
  3181. - Reuse old MTX Handle
  3182. - usbutils check all memory allocation
  3183. - usb separate thread for resource locking and modified windows locking code
  3184. - Icarus report data direction with comms errors
  3185. - Set the read and write threads for avalon to not cancel within libusb
  3186. functions and wait for the threads to pthread_join on shutdown.
  3187. - Offset needs to be incremented after avalon reads.
  3188. - Make the avalon_read function parse the ftdi responses appopriately.
  3189. - Use the avalon read timeout to completion if no data has been read.
  3190. - wait_avalon_ready should only be used before writes.
  3191. - Ask for the correct amount to read in avalon get results.
  3192. - Spawn the avalon read thread first with info->reset set to discard any data
  3193. till work is adequately queued.
  3194. - Use direct usb read commands to avoid ftdi data being automatically cut off in
  3195. avalon reads.
  3196. - Do a simple usb_read_once for the avalon result from a reset command.
  3197. - Make sure avalon is ready to receive more usb commands before sending them.
  3198. - Implement avalon_ready and avalon_wait_ready functions for when usb is ready
  3199. to receive commands.
  3200. - avalon_read should not loop but just return whatever it has succeeded in
  3201. reading.
  3202. - Set avalon_info to device data void struct.
  3203. - Specify avalon in avalon_reset.
  3204. - First pass rewriting serialdev into direct usb dev for avalon driver.
  3205. - Define a cts equivalent for direct usb and use it for avalon driver full.
  3206. - Compile usbutils into avalon driver.
  3207. - Check results come in at least at 2/3 the rate they should be on avalon and if
  3208. not, reset it.
  3209. - Give a warning but don't reset if the avalon buffer is full early.
  3210. - Discard any reads obtained from the avalon get results thread during a reset.
  3211. - Differentiate initial reset in avalon from subsequent ones.
  3212. - Perform a mandatory reset if the avalon buffer signals it's full before it has
  3213. queued its normal quota of work.
  3214. - Wait till buffer is cleared after sending idle tasks to avalon before
  3215. returning from avalon_idle.
  3216. - Lock qlock mutex during reset from read thread in avalon to prevent more work
  3217. being sent till the reset is over.
  3218. - Reset avalon if we continue to be unable to send all the work items.
  3219. - Add avalon reset response to debugging output.
  3220. - Do a wait_avalon_ready before sending a reset code.
  3221. - Iterate over spare bytes in the avalon result returned from a reset request
  3222. trying to find the beginning of the reset.
  3223. - Idle avalon after reset.
  3224. - Check for nothing but consecutive bad results on avalon and reset the FPGA if
  3225. it happens.
  3226. - Make submit_nonce return a bool for whether it's a valid share or not.
  3227. - Unset the work restart flag sooner in avalon_flush_work to avoid re-entering
  3228. the flush work function and just reset the queued counter instead of rotating
  3229. the array to avoid runs of no valid work.
  3230. - Implement an avalon_flush_work function for work restarts.
  3231. - Shut down avalon read and write threads and idle the miners on closing it.
  3232. - Tighter control over work submissions in avalon allows us to use a smaller
  3233. array.
  3234. - Rotate avalon array to reset the queued count before releasing the lock so
  3235. work will always be available on next pass.
  3236. - Move avalon read thread start till after conditional wait, store idle status
  3237. in avalon_info and use it to determine whether an error is appropriate or not.
  3238. - Wait till the avalon_send_tasks thread has filled the avalon with idle work
  3239. before starting the avalon_get_results thread.
  3240. - Use AVA_GETS_OK macro in avalon_read.
  3241. - Do all writes on avalon with a select() timeout to prevent indefinite blocking
  3242. and loop if less than desired is written.
  3243. - Check explicitly that ava_buffer_full equals the macro.
  3244. - Send initial reset as an avalon task to remove avalon_write function.
  3245. - avalon_clear_readbuf is no longer required.
  3246. - Check for 2 stray bytes on avalon reset.
  3247. - Create a separate thread for handling all work and idle submission to the
  3248. avalon which messages the scanhash function it has completed to update
  3249. statistics.
  3250. - usbutils ensure it compiles without stats
  3251. - usbutils include transfer mode in usbstats
  3252. - Give the avalon get results thread name the device number as well.
  3253. - Make sure we're not adjusting temps on every successful work retrieval on
  3254. avalon.
  3255. - Count missing work items from behind a successful work read in avalon as well.
  3256. - Change message for work not found in avalon parser.
  3257. - usbutils handle bulk_transfer partial writes
  3258. - Simplify debugging and only discard from avalon read buffer if at least one
  3259. full result has been discarded.
  3260. - Only display discarded bytes in avalon if they're not used as nonces.
  3261. - Only loop once through avalon_parse_results, but do so after timeouts as well.
  3262. - Only debug and move ram if spare bytes exist in avalon buffer.
  3263. - Remove off by one error.
  3264. - Inverted logic.
  3265. - Add more debugging to avalon reads.
  3266. - Convert unsigned size_ts to ints for parsing avalon messages.
  3267. - Cope with not finding nonces in avalon parsing gracefully by not overflowing
  3268. buffers.
  3269. - Adjust avalon temp values on one lot of valid nonces from the parser.
  3270. - Created a threaded message parser for avalon reads.
  3271. - Avalon_wait_write is not effective during resets so do it after going idle.
  3272. - Send only a single byte reset.
  3273. - Repeat going idle after avalon reset, and wait for write ready before sending
  3274. each reset request instead of some arbitrary sleep time.
  3275. - Timeouts on avalon_read and avalon_write should be 100ms.
  3276. - Don't close avalon after detecting it until we're cleaning up, instead using
  3277. reset for comms failures.
  3278. - Check for avalon_wait_write before sending reset command.
  3279. - Sleep in avalon_write_ready.
  3280. - Make avalon_wait_write a bool function and check its return value.
  3281. - Show how many idle tasks are sent to avalon if it aborts on buffer full.
  3282. - Reset avalon->device_fd after it is closed.
  3283. - Create an avalon_wait_write function that is used before sending avalon idle
  3284. command.
  3285. - Avoid repeating avalon_idle in do_avalon_close and extra sleep.
  3286. - Pass fd to avalon_idle.
  3287. - Do avalon_reset after info structure is set up.
  3288. - Rework avalon reset sequence to include idling of chips and waiting for them
  3289. to go idle followed by 2nd reset and then checking result.
  3290. - Do a non-blocking read of anything in the avalon buffer after opening the
  3291. device.
  3292. - Assign the avalon info data to the device_data in cgpu_info.
  3293. - thread shutdown is different on windows
  3294. - usbutils make all windows timeouts 999ms
  3295. - usb add another Cairnsmore1 USB chip
  3296. - icarus do the full detect test twice if required
  3297. - CMR usb config guess
  3298. - usb add transfer_read and commented out in icarus
  3299. - usbutils allow unrounded control transfers
  3300. - icarus ICA initialisation
  3301. - icarus report err on read failure
  3302. - icarus correct device_id and use device_data for icarus_info
  3303. - miner.h remove unused device_file and add device_data
  3304. - miner.h icarus no long uses fd
  3305. - icarus AMU config transfers
  3306. - Create a logwin_update function which mandatorily updates the logwin and use
  3307. it when input is expected to prevent display refresh delays.
  3308. - usbutils force an unknown IDENT for zero
  3309. - icarus set default options/timing based on device
  3310. - Must unlock curses as well in logwin_update.
  3311. - Create a logwin_update function which mandatorily updates the logwin and use
  3312. it when input is expected to prevent display refresh delays.
  3313. - icarus report usb write error information
  3314. - Add name to icarus copyright notice.
  3315. - Check for *pth dereference on pthread_join
  3316. - usbutils name latency correctly
  3317. - Check for restart before buffering more reads in Icarus.
  3318. - Icarus should timeout if it's greater than the timeout duration even if it's
  3319. receiving data.
  3320. - We should check for amount buffered in icarus get_nonce against amount already
  3321. received.
  3322. - Make mining threads report out during work submission.
  3323. - submit_work_async is no longer used directly by driver code.
  3324. - Fix first read timeout on icarus get nonce.
  3325. - Retry icarus_initialise if the first read attempt fails.
  3326. - Properly pthread_join miner threads on shutdown.
  3327. - Properly pthread_join miner threads on shutdown.
  3328. - Use a persistent single separate thread for stratum share submission that uses
  3329. workqueues since all stratum sends are serialised.
  3330. - All stratum calls to recv_line are serialised from the one place so there is
  3331. no need to use locking around recv().
  3332. - Only allow the mining thread to be cancelled when it is not within driver
  3333. code, making for cleaner shutdown and allowing us to pthread_join the miner
  3334. threads on kill_work().
  3335. - Only allow the mining thread to be cancelled when it is not within driver
  3336. code, making for cleaner shutdown and allowing us to pthread_join the miner
  3337. threads on kill_work().
  3338. - Set pool->probed to true after an attempt to resolve the url via stratum code.
  3339. - icarus test nodev everywhere
  3340. - usbutils/icarus separate FTDI transfer values and more debug
  3341. - add icarus to hotplug
  3342. - usbutils add rest of icarus
  3343. - simple serial-USB python test script
  3344. - icarus->USB v0.1 incomplete - missing initialise()
  3345. - README spelling
  3346. - Update documentation for icarus switch to USB
  3347. - Add USB rules for supported USB devices
  3348. - switch icarus configuration to usb
  3349. - usbutils new command for icarus
  3350. - usb add a numeric sub-indentity for each name
  3351. - usbutils - make FTDI handling automatic
  3352. - fix duplicate name
  3353. - usbutils set Black Arrow Lancelot's as BAL and match the lot->llt name
  3354. - usbutils identify Icarus devices
  3355. - libusb_control_transfer 16 bit words are endian specific.
  3356. - usb_applog separate amt display
  3357. - Show pool difficulty more verbosely if it changes via stratum.
  3358. - Attribute whatever stats we can get on untracked stratum shares based on
  3359. current pool diff.
  3360. - Provide a --lowmem option which does not cache shares on failed submission to
  3361. prevent low memory hardware (eg Avalon) from crashing.
  3362. - Update util.c
  3363. Version 3.1.1 - May 11th, 2013
  3364. - Use a discrete device target for scrypt that dynamically changes to ensure we
  3365. still report a work utility even if no shares are submitted such as in solo
  3366. mining.
  3367. - Make set_work_target a function to set a specified char as target for use
  3368. elsewhere.
  3369. - Further consolidate the hash regeneration between sha and scrypt doing it only
  3370. once and always checking the share diff for both before submission.
  3371. - Regenerate the hash before checking the share diff in hashtest().
  3372. - Minor typo.
  3373. - Use a scantime of 30 seconds for scrypt if none is specified.
  3374. - Support more shares to be returned for scrypt mining.
  3375. - Update the write config to properly record device entries and remove disabled
  3376. option.
  3377. - Show a different warning and loglevel for failure to resolve a URL on first or
  3378. subsequent testing of stratum pool URLs.
  3379. - Fix the problem of seting up termio of ttyUSB0 for icarus. the CSIZE is the
  3380. mask of CS2/4/8 From: navyxliu <navy.xliu@gmail.com>
  3381. - Set all stratum sockets to nonblocking to avoid trying to use MSG_DONTWAIT on
  3382. windows.
  3383. - Fix warnings on win32 build.
  3384. - Only use MSG_NOSIGNAL for !win32 since it doesn't exist on windows.
  3385. - Use MSG_NOSIGNAL on stratum send()
  3386. - Set TCP_NODELAY for !linux for raw sockets.
  3387. - Use TCP_NODELAY with raw sockets if !opt_delaynet
  3388. - Make raw sockets compile on windows
  3389. - Recheck select succeeds on EWOULDBLOCK for stratum.
  3390. - usbutils/mmq fixed size usb_read default to wait for all data
  3391. - usbutils optional (disabled by default) dev debug
  3392. - Add an ftdi usb read macro without newline
  3393. - Avalon usb interface should be 0.
  3394. - Add more debug for failure to USB init.
  3395. - Recv() should all be non-blocking for raw sockets in stratum.
  3396. - Change verbosity and error for getaddrinfo warnings in setup stratum socket.
  3397. - Free servinfo after p is checked in setup stratum socket.
  3398. - Use raw sockets without curl for stratum communications.
  3399. - Sacrifice curl handle memory on stratum disconnects on all versions of libcurl
  3400. to avoid curl corruption.
  3401. - Don't use TCP_NODELAY if opt_delaynet is enabled with stratum.
  3402. - Fix warnings in avalon driver.
  3403. - Make FULLNONCE an ULL to fix a warning on 32 bit.
  3404. - ztx correct applog typing
  3405. - ocl correct applog typing
  3406. - util correct applog typing
  3407. - api correct applog typing
  3408. - cgminer correct applog typing
  3409. - scrypt correct applog typing
  3410. - bfl correct applog typing
  3411. - ica correct applog typing
  3412. - mmq correct applog typing
  3413. - adl fix trailing %
  3414. - usbutils correct applog typing
  3415. - applog - force type checking
  3416. - Simplify the many lines passed as API data in the avalon driver now that the
  3417. API does not need persistent storage for the name.
  3418. - Duplicate the name string always in api_add_data_full to not need persistent
  3419. storage for names passed to it.
  3420. - Add extra matching work count data in API for Avalon with 4 modules.
  3421. Version 3.1.0 - April 28th, 2013
  3422. - va_copy is meant to be matched by a va_end in log_generic.
  3423. - usbutils remove_in_use break
  3424. - usbutils remove_in_use missing prev
  3425. - usbutils missing add_in_use
  3426. - Clean up summary slightly better on exit.
  3427. - Make the scan sleep time after scanwork in bflsc dynamic to keep queues
  3428. between watermark levels.
  3429. - Remove unused temp counts in bflsc.
  3430. - Calculate a rolling 5 min average set of temperatures for bflsc.
  3431. - Damp the display of voltage for BFLSC devices.
  3432. - Damp the temperature display measurement for bflsc since it fluctuates so
  3433. wildly.
  3434. - bflsc add volt stats
  3435. - Handle failed tolines command in bflsc driver.
  3436. - Can use a read lock instead of a write lock in bflsc scanwork.
  3437. - Since we are filling a queue on the bflsc devices, there is no need to run
  3438. through scanwork frequently provided we use the restart_wait function to abort
  3439. early during a block change.
  3440. - Remove flushed work in bfl scanwork from the hash table.
  3441. - Set correct device in process_nonces in bflsc driver.
  3442. - bflsc add work reply INPROCESS: missing from the spec
  3443. - bflsc put in some error messages not yet written
  3444. - bflsc get completed hashes as late as possible
  3445. - Fix potential memory leak with unused work items in bflsc_queue_full
  3446. - Reverse bools in bflsc_queue_full
  3447. - Avoid recursive loop calling correct function instead.
  3448. - bflsc fix details identification
  3449. - Differentiate BFLSC device from regular bitforce and give warning if no
  3450. support is compiled in.
  3451. - util.c str_text make a fully text readable version of str
  3452. - BFLSC fix FPGA identity overlap
  3453. - Locking error in bflsc_send_work
  3454. - Use htobe32 function for converting nonce in bflsc.
  3455. - Replace deprecated bzero with memset in bflsc driver.
  3456. - Fix compilation of bflsc driver without opencl.
  3457. - Check for realloc failures in bflsc driver.
  3458. - Check for failure to calloc in bflsc driver.
  3459. - Trivial style change
  3460. - Use copy_time function in bflsc driver.
  3461. - Use cgtime in bflsc driver and update copyright notice.
  3462. - Use a separate function for bfl initialise that doesn't require locking.
  3463. - Fix BFLSC building.
  3464. - bflsc v0.1
  3465. Version 3.0.1 - April 25th, 2013
  3466. - Bypass attempting to read and save binary files on OSX to avoid crashes on >1
  3467. GPU.
  3468. - Receive failures in recv_line should unconditionally fail.
  3469. - Use sock_blocks in api.c
  3470. - Use sock_blocks function for stratum send and receive.
  3471. - Create an OS specific sock_blocks function.
  3472. Version 3.0.0 - April 22nd, 2013
  3473. - Further fix distdir for hexdump.c
  3474. - Fix build and distdir.
  3475. - Remove all CPU mining code.
  3476. - compile on win32
  3477. - Update SCRYPT README with improved hashrates for 7970.
  3478. - Use copy_time helper throughout cgminer.c
  3479. - Provide wrappers for commonly used timer routines with API stats.
  3480. - Avoid one cgtime call in sole_hash_work.
  3481. - Fulltest is true if value is <= target.
  3482. - Use system host to endian functions for clarity in fulltest.
  3483. - Provide endian_flipX functions to avoid special casing big endian in cgminer.c
  3484. - Provide a flip128 helper to simplify big endian flipping.
  3485. - Use flip helpers to simplify code for calculation of midstate.
  3486. - Use flip32 function instead of open coding it in gen_stratum_work.
  3487. - Move util.c exports to util.h
  3488. - Fix warning on building avalon on win32
  3489. - Use cgtime in driver-avalon.c
  3490. - Use cgtime in driver-icarus.c
  3491. - Use cgtime in driver-bitforce.c
  3492. - Use cgtime in logging.c
  3493. - Use cgtime in usbutils.c
  3494. - Use cgtime in driver-opencl.c
  3495. - Use cgtime wrapper in driver-modminer.c
  3496. - Use cgtime in driver-ztex.c
  3497. - Use cgtime in compat.h
  3498. - Use cgtime instead of gettimeofday in fpgautils.c
  3499. - Replace gettimeofday usage in cgminer.c with cgtime
  3500. - Create a cgminer specific gettimeofday wrapper that is always called with tz
  3501. set to NULL and increases the resolution on windows.
  3502. - Add high resolution to nmsleep wrapper on windows.
  3503. - Set default ocl work size for scrypt to 256.
  3504. - define le32toh if needed
  3505. - fliter out the wrong result from adjust fan code
  3506. - compile avalon driver on win32 and win64
  3507. - Restart threads on the rare chance we found the block ourselves.
  3508. - Add more FAQs about crossfire.
  3509. - Set last device valid work on adding device.
  3510. - Increment last device valid work count in submit_nonce to cover scrypt.
  3511. - Set opt_scrypt drv max diff for correctness.
  3512. - Make scrypt submission use the submit_nonce code, with nonces matching
  3513. endianness.
  3514. - Do testing for HW errors on submit nonce for both scrypt and sha.
  3515. - Increment hardware error count from the one site.
  3516. - Rename scrypt regenhash function for consistency.
  3517. - Add new best share info to verbose logging.
  3518. - Add notice for when network diff is changed.
  3519. - Convert error getting device IDs in ocl code to info log level only since
  3520. multiple platforms may be installed and the error is harmless there.
  3521. - Unnecessary extra array in ocl code.
  3522. - Further driver FAQs.
  3523. - Add MAC FAQ.
  3524. - Add more FAQ details.
  3525. - Check for work restart after disable in the hash queued work loop since it may
  3526. be a long time before we re-enable a device.
  3527. - Unconditionally test for many wrong results on avalon and reset to avoid
  3528. passing a corrupt avalon result to temperature code.
  3529. - build out of source dir
  3530. - Set device_diff for queued work or there will be no diff1 share count.
  3531. - Only reset an avalon device with no results when there are no results
  3532. consecutively.
  3533. - More FAQs.
  3534. - More FAQs.
  3535. - Cleanup when stratum curl fails to initialise.
  3536. - Avoid applog in recalloc_sock.
  3537. - Avoid applog under stratum_lock in recv_line.
  3538. - Avoid applog under stratum_lock in __stratum_send.
  3539. - Put spacing around locking in util.c for clarity.
  3540. - Avoid applog under cg_wlock.
  3541. - Put spacing around locking code for clarity.
  3542. - Avoid applog under pool_lock.
  3543. - Avoid more recursive locks.
  3544. - Avoid applog while ch_lock is held.
  3545. - Avoid recursive locks in fill_queue.
  3546. - Variable is already initialised in global scope.
  3547. - More GPU faqs.
  3548. - More README faqs.
  3549. - Yet more README faqs.
  3550. - Add more faqs to README.
  3551. - Wrap result wrong tests in avalon scanhash in unlikely() and only consider a
  3552. hash count of zero wrong if a restart wasn't issued.
  3553. - avalon: if result_wrong >= get_work_count jump out the read loop
  3554. - Fix warning on 32bit.
  3555. - Fix warning on 32bit.
  3556. - Avoid curl_easy_cleanup on old curl versions in setup_stratum_curl as well.
  3557. - fix the fan control on max temp2/3
  3558. - for some reason network down. one simple cgminer command: "cgminer -o
  3559. 127.0.0.1:8888 -O fa:ke --avalon-options 115200:32:10:50:256" can idle the
  3560. avalon for safe power and protect chip
  3561. - if hash_count == 0; reinit avalon, fix the 0MHS bug use the max value of temp1
  3562. and temp2 for fan control
  3563. - Reinstate the matching_work_count per subdevice on avalon based on the work
  3564. subid.
  3565. - Avalon driver is missing the drv_id.
  3566. - Rationalise and simplify the share diff and block solve detection to a common
  3567. site.
  3568. - Rationalise and simplify the share diff and block solve detection to a common
  3569. site.
  3570. - Make the avalon array size a macro.
  3571. - Use replacement of work items in the avalon buffer as needed instead of
  3572. flushing them.
  3573. - Reinstate wrong work count to reset avalon regardless and display number of
  3574. wrong results.
  3575. - Revert "The result_wrong measurement for avalon is continually leading to
  3576. false positives so remove it."
  3577. - select() on serial usb in avalon does not work properly with zero timeout.
  3578. - The result_wrong measurement for avalon is continually leading to false
  3579. positives so remove it.
  3580. - Revert "Use only 2 queued work arrays in avalon."
  3581. - Use no timeout on further reads in avalon_gets
  3582. - Do sequential reads in avalon_get_reset to cope with partial reads.
  3583. - Show read discrepancy in avalon_get_reset.
  3584. - Reuse avalon_get_work_count variable.
  3585. - Check for AVA_GETS_RESTART when deciding if avalon has messed up.
  3586. - Make the detection of all wrong results on avalon much more conservative to
  3587. avoid false positives on work restarts.
  3588. - Show error codes on select and read fail in avalon.
  3589. - If we get a restart message in avalon_gets still check if there's a receive
  3590. message to parse first without a timeout before returning AVA_GETS_RESTART.
  3591. - Use only 2 queued work arrays in avalon.
  3592. - avalon_gets is always called from the one call site so inline it.
  3593. - The read_count is unused by the avalon get result code and no longer required
  3594. for avalon reset so simplify code removing it.
  3595. - Use a separate avalon_get_reset function for resetting avalon instead of using
  3596. avalon_get_result.
  3597. - The current hash count returned by avalon scanhash is just an obfuscated
  3598. utility counter so make it explicit.
  3599. - Check for a restart before a timeout in message parsing code in avalon.
  3600. - We should check for a restart message before checking for a timeout in avalon
  3601. scanhash.
  3602. - Store the subid for the work item in avalon.
  3603. - usbutils more stats for bflsc
  3604. - Fix record_temp_fan function in avalon driver. Patch by Xiangfu
  3605. <xiangfu@openmobilefree.net>
  3606. - Remove inappropriate memset of struct avalon result which was corrupting fan
  3607. values.
  3608. - Fix warning with no curses built in.
  3609. - Bump version to 2.11.4
  3610. - Add API support for Avalon.
  3611. - Only do_avalon_close once on multiple errors.
  3612. - Reset the result_wrong count on block change in avalon scanhash to prevent
  3613. false positives for all nonces failed.
  3614. - Small timeouts on select() instead of instant timeout increase reliability of
  3615. socket reads and writes.
  3616. - Only get extra work in fill_queue if we don't have any unqueued work in the
  3617. list.
  3618. - Small timeouts on select() instead of instant timeout increase reliability of
  3619. socket reads and writes.
  3620. - Rotate the avalon work array and free work on AVA_SEND_BUFFER_EMPTY as well.
  3621. - Only get extra work in fill_queue if we don't have any unqueued work in the
  3622. list.
  3623. - Don't get any work if our queue is already full in avalon_fill.
  3624. - Differentiate socket closed from socket error in recv_line.
  3625. - Differentiate socket closed from socket error in recv_line.
  3626. - Free avalon->works in the event we call avalon_prepare on failure to
  3627. initialise.
  3628. - Fix warnings.
  3629. - Create an array of 4 lots of work for avalon and cycle through them.
  3630. - Remove unused per unit matching work count for avalon.
  3631. - Rename the confusing avalon_info pointer.
  3632. - Simplify avalon scanhash code using the new find_queued_work_bymidstate
  3633. function. Partially works only.
  3634. - Members of cgpu_info for avalon are not meant to be in the union.
  3635. - Use correct struct device_drv for avalon_drv.
  3636. - cgminer.c -S help to only say Icarus
  3637. - Check enough work is queued before queueing more in avalon_fill.
  3638. - Actually put the work in the avalon queue.
  3639. - Rneame avalon_api to avalon_drv.
  3640. - First draft of port of avalon driver to new cgminer queued infrastructure.
  3641. - Add Makefile entry for driver-avalon.
  3642. - Add configure support for avalon.
  3643. Version 2.11.4 - April 5th, 2013
  3644. - Remove bfl-sc option from configure for 2.11 branch.
  3645. - Only update hashrate calculation with the log interval.
  3646. - Update the total_tv_end only when we show the log to prevent failure to update
  3647. logs.
  3648. - Minor README updates.
  3649. - Add example 7970 tuning for scrypt in readme.
  3650. - Update driver recommendations.
  3651. - Add extensive GPU FAQs for the flood of new Scrypt miners.
  3652. - Remove help option for cpumining in build environment.
  3653. - Remove scripts that make it too easy to compile CPU mining support.
  3654. - Win32 and win64 build updates
  3655. - Remove references to CPU mining from README.
  3656. - Show share hash as little endian as needed.
  3657. - usbutils extra message requirements
  3658. - Make hashmeter frequency for hash_queued_work match sole_work.
  3659. - Update links and recommended SDKs.
  3660. - Update scrypt readme re drivers and sdk.
  3661. - usbutils.c usb_cmdname() usb_cmds -> string name
  3662. - BFL FPGA Windows timeout set to 999ms
  3663. - AUTHORS - spam update time (one year since the last)
  3664. - Update README for x970 memdiff values.
  3665. - Update README to match changes to display.
  3666. - Remove increasingly irrelevant discarded work from status lines.
  3667. - Remove increasingly irrelevant queued and efficiency values from status and
  3668. move WU to status line.
  3669. - Allow cgminer to start if usb hotplug is enabled but no devices yet exist.
  3670. - Do not scan other gpu platforms if one is specified.
  3671. - Update README for sync objects on windows.
  3672. - Update README about intensity.
  3673. - Add information for setting gpu max alloc and sync parameters for windows with
  3674. scrypt.
  3675. - If the hashmeter is less than the log interval and being updated by the
  3676. watchdog, don't update the hashrate.
  3677. Version 2.11.3 - March 17, 2013
  3678. - Update docs and reorder README to show executive summary near top.
  3679. - Update the hashmeter at most 5 times per second.
  3680. - Usbutils use its own internal read buffer
  3681. - Calculate work utility for devices that support target diffs of greater than
  3682. 1, and update scrypt code to use it.
  3683. - usbutils allow read termination match to be a string
  3684. - Set default GPU threads to 1 for scrypt.
  3685. - Connect backup stratum pools if the primary pool cannot deliver work.
  3686. - Use a new algorithm for choosing a thread concurrency when none or no shader
  3687. value is specified for scrypt.
  3688. - Do not round up the bufsize to the maximum allocable with scrypt.
  3689. - Remove the rounding-up of the scrypt padbuffer which was not effectual and
  3690. counter-productive on devices with lots of ram, limiting thread concurrencies
  3691. and intensities.
  3692. - bufsize is an unsigned integer, make it so for debug.
  3693. - Update the hashmeter once per second but only display the extra logs every
  3694. opt_log_inteval.
  3695. - add a dummy ztex to usbutils so cgminer -n lists ztex also
  3696. - nDevs required for -n with usb
  3697. - USB device list - convert some common error numbers to messages
  3698. - USB -n 'known' text only without ---usb-list-all
  3699. - USB modify -n and --usb-dump to only show known devices or use new
  3700. --usb-list-all option to see all
  3701. - Make pool adding while running asynchronous, using the pool test thread
  3702. functionality.
  3703. - Only curl easy cleanup a stratum curl if it exists.
  3704. - Sacrifice the ram of curl handles in stratum disconnects when we have built
  3705. with old libcurl to avoid crashes.
  3706. - cgminer -n to include a USB device list
  3707. - usbutils allow call of usb_all() from other code
  3708. - Convert gbt_lock to a cg_lock.
  3709. - Add intermediate variants of cglocks that can be up or downgraded to read or
  3710. write locks and use them for stratum work generation.
  3711. - Move the stratum and GBT data to be protected under a new cg_lock data_lock.
  3712. - Convert the ch_lock to cg_lock.
  3713. - Convert the control_lock to a cg_lock.
  3714. - Remove unused qd_lock.
  3715. - Implement cg_lock write biased rwlocks.
  3716. - do usb_initialise() after the started message so we see it
  3717. - --usb-dump display brief dump if value = 0
  3718. - USB add --usb options to limit USB device selection v0.1
  3719. Version 2.11.2 - March 9, 2013
  3720. - Whitelist AMD APP SDK 2.8 for diablo kernel.
  3721. - Cope with the highest opencl platform not having usable devices.
  3722. - Fix memory leak with share submission on GPU work structures as discovered by
  3723. twobitcoins.
  3724. - usb_cleanup() without locking.
  3725. - Use curl_easy_cleanup to close any open stratum sockets.
  3726. - Show pool number in switch message
  3727. - Don't start testing any pools with the watchpool thread if any of the test
  3728. threads are still active.
  3729. - Set sockd to false should curl setup fail on stratum.
  3730. - Close any open sockets when reusing a curl handle and reopen the socket
  3731. whenever we're retrying stratum.
  3732. - Set pool died on failed testing to allow idle flag and time to be set.
  3733. - Remove unused pthread_t typedefs from struct pool.
  3734. - Perform pool_resus on all pools that are found alive with the test pool
  3735. threads.
  3736. - Use pool_unworkable in select_balanced as well.
  3737. - Differentiate pool_unusable from pool_unworkable.
  3738. - Keep a connection open on higher priority stratum pools to fail back to them.
  3739. - Rename threads according to what pool they're associated with as well.
  3740. - Set the wrong bool in pool_active
  3741. - Start the stratum thread only if we successfully init and authorise it,
  3742. otherwise unset the init flag.
  3743. - Make the initialisation of the stratum thread more robust allowing the
  3744. watchpool thread safe access to it after the stratum thread is started.
  3745. - API no longer ignore send() status
  3746. - API make the main socket non-static
  3747. Version 2.11.1 - March 7, 2013
  3748. - Shorten the time before keepalive probes are sent out and how frequently
  3749. they're sent with stratum curls.
  3750. - Only set stratum auth once to prevent multiple threads being started.
  3751. - Display select return value on select fail in stratum thread.
  3752. - Clear the socket of anything in the receive buffer if we're going to retry
  3753. connecting.
  3754. - Allow pools to be resuscitated on first startup by the watchpool thread.
  3755. - Check all pools simultaneously at startup switching to the first alive one to
  3756. speed up startup.
  3757. - Clear just the socket buffer when we don't care what is left in a stratum
  3758. socket.
  3759. - Clear the stratum socket whenever we are closing it since the buffer is going
  3760. to be reused.
  3761. - Do not continue work from a stratum pool where the connection has been
  3762. interrupted.
  3763. - Reset stratum_notify flag on suspend_stratum as well.
  3764. - Close any sockets opened if we fail to initiate stratum but have opened the
  3765. socket.
  3766. - Close any existing stratum socket if we are attempting to restart stratum so
  3767. the pool knows the connection has gone.
  3768. - Show mechanism of stratum interruption if select times out.
  3769. - Make stratum connection interrupted message higher priority to be visible at
  3770. normal logging levels.
  3771. - Implement client.show_message support for stratum.
  3772. - API add 'Network Difficulty' to 'coin'
  3773. - Setup BFLSC support
  3774. - API use control_lock when switching pools
  3775. - Make sure to retry only once with noresume support for stratum.
  3776. - Instead of keeping track of when the last work item was generated to keep
  3777. stratum connections open, keep them open if any shares have been submitted
  3778. awaiting a response.
  3779. - usbutils.c copy full size to 'Last Command'
  3780. - configure - set USE_USBUTILS when usbutils is required and use it in the code
  3781. - Clear last pool work on switching pools if the current pool supports local
  3782. work generation or we are in failover only mode.
  3783. - make rw locks: mining_thr_lock and devices_lock
  3784. - Release MMQ device only once (not 4 times)
  3785. - api.c fix MSG overlap
  3786. - Hotplug - allow setting interval via --hotplug or API
  3787. - curses - fix - put a dev_width inside #ifdef
  3788. - usb_cleanup() use correct locking mechanism
  3789. - Implement and use usb_cleanup() on shutdown or restart
  3790. - miner.php report 'Last Valid Work' as time before request
  3791. - API - return Last Valid Work
  3792. - api -> drv
  3793. - ZTX bug set missing drv_id
  3794. Version 2.11.0 - March 2, 2013
  3795. - Update kernel file names signifying changes.
  3796. - Update a pool's last work time when the work is popped as well as staged.
  3797. - API always report failed send() replies
  3798. - Update diff stale: total and pools when stratum throws away shares
  3799. - Keep stratum connections open for 2 minutes after the last work item was
  3800. staged to allow stray shares to be submitted on pool switching.
  3801. - Try to extract the sessionid associated with mining.notify on 3rd level array
  3802. and submit it along with the userid to support mining resume, failing gracefully
  3803. and restarting if the pool rejects it.
  3804. - Speed up watchdog interval and therefore display updates to 2 seconds.
  3805. - Update copyright dates.
  3806. - Cope with misread sessionid on stratum for now.
  3807. - Use constants from the array of __constants throughout the diablo kernel.
  3808. - Create a __constant array for use within diablo kernel.
  3809. - Fix --benchmark generating valid work for cgminer.
  3810. - Use the sessionid as passed on stratum connect to attempt to resume a
  3811. connection once and then clear it if it fails, to use a new connection.
  3812. - Move to storing the nonce1 in the work struct instead of the sessionid for the
  3813. now defunct first draft mining.resume protocol.
  3814. - Use global constant arrays for all other constants used in scrypt kernel.
  3815. - Use global __constants for sha functions in scrypt kernel.
  3816. - Use constants for endian swap macros.
  3817. - Revise scrypt kernel copyright notice.
  3818. - Separate out additions in scrypt kernel.
  3819. - Reuse some Vals[] variables that can be assigned to constants earlier in the
  3820. poclbm kernel, making for fewer ops.
  3821. - Put all constants used in poclbm kernel into __const memory array to speed up
  3822. concurrent reads on the wavefront.
  3823. - BFL stop 1st init command if no device
  3824. - Add a get_queued function for devices to use to retrieve work items from the
  3825. queued hashtable.
  3826. - Bugfix: Duplicate stratum sessionid when copying work, to avoid double-free
  3827. - Bugfix: Missing pool_no parameter to applog for no-stratum-sessionid debug
  3828. message
  3829. - Add the choice of hash loop to the device driver, defaulting to hash_sole_work
  3830. if none is specified.
  3831. - Add comments.
  3832. - Add a driver specific flush_work for queued devices that may have work items
  3833. already queued to abort working on them on the device and discard them.
  3834. - Flush queued work on a restart from the hash database and discard the work
  3835. structs.
  3836. - Create a central point for removal of work items completed by queued device
  3837. drivers.
  3838. - Create a fill_queue function that creates hashtables of as many work items as
  3839. is required by the device driver till it flags the queue full.
  3840. - Create the hash queued work variant for use with devices that are fast enough
  3841. to require a queue.
  3842. - Update copyright year.
  3843. - Fix tv_lastupdate being made into tv_end and update the hashmeter on cycle,
  3844. not opt_log_interval.
  3845. - Fix tv_lastupdate being made into tv_end and update the hashmeter on cycle,
  3846. not opt_log_interval.
  3847. - Only continue submitting shares with mining.resume support on stratum when the
  3848. session id matches.
  3849. - Provide support for mining.resume with stratum, currently re-authorising after
  3850. successful resumption pending finalising of the protocol process.
  3851. - Provide basic framework for restarting stratum depending on whether resume
  3852. support exists or not.
  3853. - Abstract out the setting up of the stratum curl socket.
  3854. - Free sessionid in clean_work and remove redundant setting of strings to NULL
  3855. since the whole work struct is zeroed.
  3856. - Only clear stratum shares mandatorily on stratum dropouts when the pool does
  3857. not support resume.
  3858. - Try resubmitting stratum shares every 5 seconds for up to 2 minutes if the
  3859. pool session id exists and matches on failure to submit.
  3860. - Do as much outside of mutex locking of sshare_lock as possible.
  3861. - Remove last reference to struct work used outside the sshare_lock in
  3862. submit_work_thread
  3863. - Unlock the sshare_lock in submit_work_thread when all references to work and
  3864. sshare are complete.
  3865. - Add timestamps to stratum_share structs as they're generated and copy the
  3866. stratum sessionid if it exists to stratum work generated.
  3867. - Store session id for stratum if the pool supports it for future mining.resume
  3868. support.
  3869. - API.java allow partial reads
  3870. - debug_cb buffer type warning
  3871. - MMQ rewrite the last of the old scanhash loop and drastically reduce CPU
  3872. - hash_sole_work can be static
  3873. - Make the numbuf larger to accept larger scrypt parameters.
  3874. - Keep the unique id of each work item across copy_work to prevent multiple work
  3875. items having the same id.
  3876. - Abstract out the main hashing loop to allow us to use a separate loop for
  3877. devices that are fast enough to require queued work.
  3878. - Provide a noop thread_enable function for drivers that don't support it.
  3879. - Provide a noop thread_shutdown function for drivers that don't support it.
  3880. - Provide a noop hw_error function for drivers that don't support it.
  3881. - Provide a noop prepare_work for drivers that don't support it.
  3882. - Provide a noop thread_init for drivers that don't support it.
  3883. - Provide a noop can_limit_work for devices that don't support it.
  3884. - Provide a noop thread_prepare function for drivers that don't use
  3885. thread_prepare.
  3886. - Use blank_get_statline_before for GPU devices that don't support adl
  3887. monitoring.
  3888. - Provide a noop get_stats function for drivers that don't support it.
  3889. - Provide a blank get_statline for drivers that don't support it.
  3890. - Provide a blank get_statline_before function for drivers that don't have one.
  3891. - Fill drivers missing reinit_device with a noop version.
  3892. - add 'count' to cumstomsummarypage 'calc'
  3893. - hotplug use get_thread() where appropriate
  3894. - convert sleep(const) to nmsleep()
  3895. - remove empty #ifdef
  3896. - call a separate get_devices() with locking, as required
  3897. - usbutils - avoid free cgusb twice
  3898. - usbutils hotplug v0.1
  3899. - Report USB nodev as ZOMBIE on the screen
  3900. - Change file modes.
  3901. Version 2.10.5 - February 7, 2013
  3902. - Fix logic fail on partial writes with stratum send that was leading to corrupt
  3903. message submissions.
  3904. - Do not consider every call to stratum_resumed a pool recovery unless it was
  3905. actually idle.
  3906. - Do not enable the pool disable on reject feature unless explicitly enabled
  3907. with --disable-rejecting.
  3908. - Stratum disconnect shares - count total against stale
  3909. - Use sanity checking to prevent a possible overflow with invalid data being
  3910. given by the pool for difficulty as reported by luke-Jr.
  3911. - Check for calloc failure for completeness in gen_stratum_work.
  3912. - Cache the coinbase length to speed up stratum work generation.
  3913. - Cache the header length when generating stratum work to avoid calculating it
  3914. on every work generation, and to only need one alloc+sprintf, speeding up work
  3915. generation.
  3916. - Use heap ram for coinbase in gen_stratum_work, zeroing it before use.
  3917. - Provide a wrapper for aligning lengths of size_t to 4 byte boundaries.
  3918. - Fix memory leak on stratum share submission.
  3919. - Zero the best share string memory when zeroing stats.
  3920. Version 2.10.4 - December 29, 2012
  3921. - Change the pool stratum socket buffer to be dynamically allocated to
  3922. accomodate any size coinbase and keep receiving data in recv line for up to 60s
  3923. if no end of line has been received.
  3924. - Differentiate socket full from sock full.
  3925. - Allow stratum to startup without notify but check it is valid before creating
  3926. stratum work.
  3927. - Do not try to generate stratum work unless the notify command has succeeded.
  3928. - Reset total diff1 shares when zeroing stats as well to show correct work
  3929. utility.
  3930. Version 2.10.3 - December 26, 2012
  3931. - Do not give the share submission failure message on planned stratum
  3932. disconnects.
  3933. - Parse anything in the stratum socket if it's full without waiting. Empty the
  3934. socket even if a connection is not needed in case there are share returns.
  3935. - Provide a mechanism to zero all the statistics from the menu.
  3936. - Display the current pool diff in the status line.
  3937. - Display block diff in status line.
  3938. - Generalise the code for solving a block to enable block solve detection with
  3939. scrypt mining.
  3940. - Generate the output hash for scrypt as well and use the one function to set
  3941. share_diff.
  3942. - Use the flip80 function in regeneratehash and the correct sized hash array.
  3943. - Use one size for scratchbuf as a macro in scrypt.c
  3944. - Stage work outside of the stgd lock to prevent attempted recursive locking in
  3945. clone_available.
  3946. - share_diff needs to be performed on a BE version of the output hash to work,
  3947. leading to false best_share values as spotted by luke-Jr.
  3948. - Remove the unused sha224 functions.
  3949. - Use the flip functions in hashtest.
  3950. - Simplify the setting of the nonce data field in work on submitting nonces.
  3951. - Scrypt code does not enter the hashtest function.
  3952. - Go back to cloning available work under staged lock.
  3953. - Updated links to AMD APP SDK
  3954. - Updated link to ADL SDK
  3955. - scrypt_diff uses a uint64_t as well.
  3956. - Correct target for stratum support with scrypt mining.
  3957. - libztex: fixed a typo
  3958. - libztex: check returnvalue of libusb_claim_interface() and release the
  3959. interface in case of early exit
  3960. Version 2.10.2 - December 19, 2012
  3961. - Stop all work from the current pool if it's a stratum pool once it is
  3962. disconnected since it will be invalid upon reconnecting.
  3963. - Discard all staged work from stratum pools as well as the shares upon
  3964. disconnection since all the work becomes invalid.
  3965. - Use correct cbreak after 15 second delay when no pool is found alive.
  3966. - MMQ missing firmware -> ERR not DEBUG
  3967. - Allow stratum to work with scrypt.
  3968. - MMQ ensure delta clock can never exceed limits
  3969. - MMQ lowercase new string constants
  3970. - MMQ add api pgaset for clock
  3971. - API V1.23 - new pgaset command, to be used soon
  3972. - Protect the best_share/best_diff values under control lock.
  3973. - MMQ style police
  3974. - MMQ count work check timeout failures
  3975. - MMQ allow partial work replies and count them
  3976. - Check a stratum pool hasn't gone dead while being a backup pool and missed
  3977. having its idle flag cleared.
  3978. - MMQ overheat: remove clockdown (doesn't help) + ensure no lost shares
  3979. - API-README grammar
  3980. - API-README explain custom page extensions in miner.php
  3981. - miner.php add a sample group pool report
  3982. - miner.php allow where,group,having on cumstom pages
  3983. Version 2.10.1 - December 14, 2012
  3984. - Check for EWOULDBLOCK when supported in send and recv as well.
  3985. - Use the raw send() command instead of curl_easy_send since curl raw socket
  3986. usage introduces random bugs on windows.
  3987. - Use raw recv() command in place of curl_easy_recv since the curl
  3988. implementation introduces random bugs on windows builds when the recv fails.
  3989. - miner.php when displaying a single rig, add prev/next rig buttons if they
  3990. exist, next to refresh
  3991. - miner.php allow custom page joins for STATS
  3992. - API show if pool has GBT (so people know not to use that pool)
  3993. - miner.php - include windows easyphp link
  3994. - driver-ztex: use the correct size for the swap array
  3995. - API stats - display pool byte transfer stats
  3996. - Pool store data transfer stats
  3997. - README ModMiner dependency
  3998. - Benchmark incorrect work size
  3999. - ChangeLog refer to NEWS
  4000. - MMQ handle over temp differently and hash longer
  4001. - driver-ztex: search the complete noncerange based on the actual speed
  4002. - README - update ModMiner details
  4003. - API-README update
  4004. - api use a dynamic io buffer, truncated before it reaches the current ~64k
  4005. limit
  4006. Version 2.10.0 - December 10, 2012
  4007. - Include prctl header for thread renaming to work.
  4008. - Set tv_idle time if a pool is not active when input from the menu.
  4009. - usb display message when device is in use/another cgminer
  4010. - libztex: avoid the use of libusb_error_name()
  4011. - minor unlikely zero pointer test
  4012. - BeaverCreek doesn't like BFI INT patching.
  4013. - Only stratum pools that are idle need to be kicked via cnx_needed.
  4014. - mmq - abbreviate the temperature numbers
  4015. - Do not do any setup if opt_api_listen is disabled in api.c.
  4016. - usbutils.c uninitialised usbstat for non-primary mmqs
  4017. - Only set the lagging flag for select_pool() on failed getwork if we're not in
  4018. opt_fail_only mode.
  4019. - libztex: in case the selectFpga() failed set the selected fpga to unknown
  4020. - Modified windows-build.txt to update git instructions.
  4021. - libztex: use a function for the twice called firmware reset code
  4022. - libztex: removed an unused struct member (ztex->valid)
  4023. - driver-ztex: support for broken fpga on a multifpga board
  4024. - Set the pool lagging flag on startup to avoid it being shown initially, and
  4025. only unset it once the maximum number of staged work items has been reached.
  4026. - Avoid recursive locking of the stgd lock.
  4027. - Return value of keep_sockalive is no longer used.
  4028. - Remove dependency on mstcpip.h for windows build by making curl version >=
  4029. 7.25.0 mandatory on windows builds, and use curl functions for keepalive
  4030. whenever possible instead.
  4031. - Make main() the getwork scheduler once everything is set up, so that all app
  4032. exits use the kill_work and quit paths.
  4033. - ztex: more style and whitespace fixes
  4034. - libztex: silenced another warning
  4035. - Set successful connect to true on auth stratum to allow summary on exit from
  4036. single stratum pool.
  4037. - Only consider work stale for stratum of different job_id if it's not a share.
  4038. - Increment version preempting changed version signifying different codebase to
  4039. 2.9
  4040. - Hash_pop should signal further waiters on its own pthread conditional in case
  4041. there are multiple waiters.
  4042. - Check the job_id has not changed on stratum work when deciding if the work is
  4043. stale as might occur across disconnections.
  4044. - Perform pool_resus on getwork pool that generates work in getwork_thread.
  4045. - Set pool lagging message for getwork pool that falls to zero staged in getwork
  4046. thread.
  4047. - Stage extra work when the primary pool is a getwork pool without rolltime.
  4048. - Do not try to clean up twice if kill message is given.
  4049. - Only recalculate total_staged in getwork thread if required.
  4050. - Include the correct config header in libztex and include it before other
  4051. includes.
  4052. - Implement a completely new getwork scheduler. Stage all work from the one
  4053. thread, making it possible to serialise all requests minimising the number of
  4054. getworks requested or local work generated. Use a pthread conditional to wake up
  4055. the thread whenever work is removed to generate enough work to stay above the
  4056. watermark set by opt_queue. Remove all remnants of the old queueing mechanism,
  4057. deleting the now defunct queued count.
  4058. - libztex: fixed some warnings and removed some whitespaces
  4059. - libztex: silenced some warnings
  4060. - Remove all references to the now unused workio_cmd structure.
  4061. - Remove the old workio command queue thread, replacing it with a kill
  4062. conditional to exit the program.
  4063. - Remove getwork command from workio_cmd queues and do them directly from
  4064. queue_request.
  4065. - Begin tearing down the old workio command queues by removing submit commands
  4066. from there and submit them asynchronously via their own threads.
  4067. - Update windows build instructions.
  4068. - Set pool probed to true on successful authorisation with stratum to avoid it
  4069. being pinged later with pool_getswork.
  4070. - driver-ztex: libztex_setFreq() must be called before ztex_releaseFpga()
  4071. - driver-ztex: changed two pairs of malloc()/memset() to calloc()
  4072. - libztex: Read bitstream file in 2kb blocks with simpler and faster code
  4073. - Added the binary versions of ztex_ufm1_15d4.ihx and ztex_ufm1_15y1.ihx
  4074. - Trivial space removal.
  4075. - libztex: Add firmware download support for ZTEX 1.15d and 1.15x
  4076. - libztex: Factor out local version of libusb_get_string_descriptor_ascii()
  4077. - Shut up some boring old cpu warnings.
  4078. - Style changes.
  4079. - Allow pool active to be called on stratum or disabled pools in the watchpool
  4080. thread if the pool has not been probed.
  4081. - libztex: Make log messages say bitstream when refering to bitstreams
  4082. - libztex: Don't return error when a bitstream was already configured
  4083. - libztex: Read bitstream file in 64kb blocks with simpler and faster code
  4084. - libztex: Verify that the mining firmware is not a dummy firmware
  4085. - libztex: Match mining firmware ZTEX descriptor against the dummy firmware
  4086. - Combine shared padding into one char.
  4087. - libztex: Start download sequence only after reading in the new firmware
  4088. - libztex: Download mining firmware to all devices with dummy firmware
  4089. - lock (most of) the threaded statistics updates
  4090. - README stats don't add up
  4091. - usbutils.c remove compiler warning
  4092. - Make need connection return true if a pool is idle.
  4093. - API add Best Share to summary
  4094. - Check on creating new GBT work if the structures are up to date and update
  4095. them as required rather than regularly.
  4096. - Update windows build instructions.
  4097. - Enable backup stratum connections for getwork when the primary pool doesn't
  4098. have longpoll aka solo mining.
  4099. - Check for correct absence of opt_fail_only in cnx_needed.
  4100. - Remove unused variable.
  4101. - The specification for stratum has been elaborated to say that a changed diff
  4102. applies only to new work so do not retarget when submitting shares.
  4103. - Use a variable length string array in submit_upstream_work to cope with
  4104. massive GBT submissions.
  4105. - API lock access to some summary statistics (and copy them)
  4106. - Suspend stratum connections to backup pools when there is no requirement to
  4107. potentially grab work from them.
  4108. - Fix missing export for RenameThread.
  4109. - enumerate the mining threadnames
  4110. - MMQ avoid possible number overrun crashes
  4111. - mmq usb v0.4 + api usb stats
  4112. - setting the name of the threads for linux,freebsd,openbsd and osx code is
  4113. borrowed from bitcoins util.c, so it is already tested
  4114. - Don't show broken WU value with scrypt mining.
  4115. - Style police.
  4116. - Remove unused getwork times in getswork.
  4117. - Fix readme wordwrap.
  4118. Version 2.9.6 - December 2, 2012
  4119. - Make gen_stratum_work more robust by using a dynamically allocated array for
  4120. the header in case bogus data is sent by the pool to avoid overflowing a static
  4121. array.
  4122. - scrypt_diff now returns a uint64_t
  4123. - Support monitoring and reporting much higher diffs for scrypt mining,
  4124. truncating irrelevant zeroes from displayed hash.
  4125. - Pass ostate values around in scrypt to be able to extract full hashes if
  4126. needed later on.
  4127. - Since we will be using calloc_str to put a string into it, convert the
  4128. function to calloc_strcat which does it automatically.
  4129. - Revert "Handle crash exceptions by trying to restart cgminer unless the
  4130. --no-restart option is used."
  4131. - Count longpoll and GBT decodes as queued work since the count otherwise
  4132. remains static.
  4133. - Use the string helper functions to create gbt blocks of any length.
  4134. - Provide helper functions calloc_str and realloc_strcat to create and extend
  4135. arbitrary length arrays based on string length.
  4136. Version 2.9.5 - November 25, 2012
  4137. - fixes target calc for mips openwrt
  4138. - openwrt needs roundl
  4139. - Get rid of unused last_work in opencl thread data.
  4140. - Do away with the flaky free_work api in the driver code which would often lose
  4141. the work data in opencl and simply flush it before exiting the opencl scanhash.
  4142. - Use base_work for comparison just for cleanness in __copy_work
  4143. - Remove all static work structs, using the make and free functions.
  4144. - Add pool no. to stale share detected message.
  4145. - Add info about which pool share became stale while resubmitting.
  4146. -b Copy the work on opencl_free_work
  4147. - Add an extra slot in the max backlog for ztex to minimise dupes.
  4148. - Do not use or count the getworks submitted which are simply testing that pools
  4149. are still up. This was increasing share leakage and making stats not reflect
  4150. real work.
  4151. - Track all dynamically allocated memory within the work struct by copying work
  4152. structs in a common place, creating freshly allocated heap ram for all arrays
  4153. within the copied struct. Clear all work structs from the same place to ensure
  4154. memory does not leak from arrays within the struct. Convert the gbt coinbase and
  4155. stratum strings within the work struct to heap ram. This will allow arbitrary
  4156. lengths without an upper limit for the strings, preventing the overflows that
  4157. happen with GBT.
  4158. - libztex: Work around ZTEX USB firmware bug exposed by the FreeBSD libusb
  4159. - opencl: Use new dev_error function for REASON_DEV_NOSTART
  4160. Version 2.9.4 - November 18, 2012
  4161. - Provide rudimentary support for the balancing failover strategies with stratum
  4162. and GBT by switching pools silently on getwork requests.
  4163. - Convert remaining modminer and bfl uses of usleep to nmsleep.
  4164. - Convert libztex to nmsleep where possible.
  4165. - Convert unreliable usleep calls to nmsleep calls in ztex driver.
  4166. - Support workid for block submission on GBT pools that use it.
  4167. - Provide rudimentary support for literal ipv6 addresses when parsing stratum
  4168. URLs.
  4169. - Work around libcurl cflags not working on hacked up mingw installations on
  4170. windows.
  4171. - Only increase gpu engine speed by a larger step if the temperature is below
  4172. hysteresis instead of increasing it to max speed.
  4173. - Convert pool not responding and pool alive message on backup pools to verbose
  4174. level only since they mean a single failed getwork.
  4175. - Update work block on the longpoll work item before calling restart threads to
  4176. ensure all work but the longpoll work item gets discarded when we call
  4177. discard_stale from restart_threads.
  4178. - Do not attempt to remove the stratum share hash after unsuccessful submission
  4179. since it may already be removed by clear_stratum_shares.
  4180. - Check against a double for current pool diff.
  4181. - Support for fractional diffs and the classic just-below-1 share all FFs diff
  4182. target.
  4183. Version 2.9.3 - November 11, 2012
  4184. - Make header larger on gen stratum work to accomodate \0 at the end.
  4185. Version 2.9.2 - November 11, 2012
  4186. - Use stratum block change from backup pools as an alternative to longpoll for
  4187. pools that don't support LP.
  4188. - Check share target diff for best_share to be calculated when solo mining.
  4189. - Round some more static string arrays to 4 byte boundaries.
  4190. - There is no need for the static arrays to be larger than required, so long as
  4191. they're 4 byte aligned to appease ARM.
  4192. - Store the full stratum url information in rpc_url for correct configuration
  4193. file saving.
  4194. - Put in a hack to prevent dud work from sneaking into test_work_current being
  4195. seen as a new block.
  4196. - Reset the work->longpoll flag where it will affect stratum work items as well.
  4197. - Check for both coinbase/append and submit/coinbase support before using GBT
  4198. protocol.
  4199. - First pass through testing for GBT should not set probed to true since we are
  4200. about to probe again.
  4201. - Hash1 is only used by the deprecated cpu mining code and never changes so
  4202. remove it from the work struct and bypass needing to process the value for all
  4203. other mining.
  4204. - Get a work item once per minute for all getwork and GBT pools to test they're
  4205. still alive and to maintain a current GBT template.
  4206. - Get a fresh block template with GBT pools on switching to them.
  4207. Version 2.9.1 - November 6, 2012
  4208. - Reset work flags to prevent GBT shares from being submitted as stratum ones
  4209. after switching.
  4210. Version 2.9.0 - November 6, 2012
  4211. - Add endian swap defines for where missing.
  4212. - Only retarget stratum shares to new pool diff if diff has dropped.
  4213. - Remove resetting of probed variable when detecting GBT.
  4214. - Count lost stratum share submits and increase message priority to warning.
  4215. - Only retrieve a new block template for GBT pools that are the current pool.
  4216. - Show which pool untracked share messages have come from.
  4217. - Add management for dead GBT pools.
  4218. - Count lost shares with stratum as submit stale lost.
  4219. - Discard record of stratum shares sent and report lost shares on disconnection
  4220. since they will never be reported back.
  4221. - Swab, don't just swap the bytes in the GBT target.
  4222. - Change status window message for GBT connected pools versus LP.
  4223. - Generate a gbt work item from longpoll when required to set new block and
  4224. message appropriately.
  4225. - Use existing pool submit_old bool from gbt data.
  4226. - Retrieve a new block template if more than 30 seconds has elapsed since the
  4227. last one to keep the data current and test the pool is still alive.
  4228. - Update GBT longpollid every time we request a new longpoll.
  4229. - Manage appropriate response codes for share submission with GBT.
  4230. - Allow the longpoll thread to start with GBT and only set the longpollid once.
  4231. - Correct last few components of GBT block generation courtesy of Luke-jr.
  4232. - Use correct length for offsetting extra nonce and remaining data.
  4233. - Flip all 80 bytes in the flip function which was wrongly named flip256 for its
  4234. purpose.
  4235. - Calculate midstate for gbt work and remove now unused variable.
  4236. - Use a standard function for flipping bytes.
  4237. - Insert the extra nonce and remaining data in the correct position in the
  4238. coinbase.
  4239. - Remove txn size debugging and enlarge gbt block string to prevent overflow.
  4240. - Remove varint display debugging.
  4241. - Build varint correctly for share submission and sleep 5 seconds before
  4242. retrying submit.
  4243. - Make gbt_coinbase large enough for submissions, swap bytes correctly to make a
  4244. header from GBT and encode the number of transactions in share submission.
  4245. - Store the fixed size entries as static variables in GBT in binary form,
  4246. byteswapping as is required.
  4247. - 32 bit hex encoded variables should be in LE with GBT.
  4248. - Target and prevblockhash need to be reversed from GBT variables.
  4249. - Construct block for submission when using GBT.
  4250. - Use same string for debug as for submission and make string larger to cope
  4251. with future GBT messages.
  4252. - Skip trying to decipher LP url if we have GBT support.
  4253. - Store all the transaction hashes in pool->txn_hashes instead of separating
  4254. txn0 and correct generation of merkle root, fixing memory overwrites.
  4255. - Hook into various places to generate GBT work where appropriate.
  4256. - Create extra work fields when generating GBT work.
  4257. - Generate header from correct hashing generation of the merkle root for GBT.
  4258. - Generate the merkle root for gbt work generation.
  4259. - Create a store of the transactions with GBT in the minimum size form required
  4260. to generate work items with a varied coinbase.
  4261. - Create a function that generates a GBT coinbase from the existing pool
  4262. variables.
  4263. - Extract and store the various variables GBT uses when decoding gbt work.
  4264. - Check for invalid json result in work_decode.
  4265. - Decode work in separate functions for getwork vs gbt.
  4266. - Check for the coinbase/append mutable in GBT support to decide whether to use
  4267. it or not.
  4268. - Add a gbt mutex within the pool struct for protecting the gbt values.
  4269. - Convert work decode function to prepare for decoding block templates.
  4270. - Check for GBT support on first probing the pool and convert to using the GBT
  4271. request as the rpc request for that pool.
  4272. - Make the rpc request used with getwork a pool variable to allow it to be
  4273. converted to/from gbt requests.
  4274. - Changes to build prototypes to support building on FreeBSD 9.1-RC2 amd64
  4275. - Free old stratum_work data before replacing it
  4276. - There is no need for addrinfo any more.
  4277. - server and client sockaddr_in are no longer used in struct pool.
  4278. - Merge pull request #322 from luke-jr/bugfix_stratum_tmpwork
  4279. - Set sshare id and swork_id within the sshare mutex to avoid multiple share
  4280. submits with the same id.
  4281. - Initialize temporary stratum work
  4282. Version 2.8.7 - October 29, 2012
  4283. - Fail on select() failing in stratum thread without needing to attempt
  4284. recv_line.
  4285. - Add share to stratum database before sending it again in case we get a
  4286. response from the pool before it's added.
  4287. Version 2.8.6 - October 29, 2012
  4288. - Shorten the initiate stratum connect timeout to 30 seconds.
  4289. - Shorten the stratum timeout on read to 90 seconds to detect unresponsive pool.
  4290. - Display best share difficulty on exit.
  4291. - Make stratum socket fail more robust on windows by disabling the send buffer.
  4292. - Reuse the same curl handle forcing a new connection instead of risking
  4293. derefencing.
  4294. - Add information about submission failure to stratum send.
  4295. - Only add stratum share to database if we succeeded in submitting it, with a
  4296. debug output saying it succeeded.
  4297. - Use keepalive with stratum sockets to improve its ability to detect broken
  4298. connections.
  4299. - Show only the URL in the status bar to avoid long prefixes making for extra
  4300. long lines.
  4301. - Display compact status in menu and update README to reflect current menu
  4302. entries.
  4303. - Add a compact display mode that does not list per device statistics in the
  4304. status window.
  4305. - Add blank spaces after best share displayed.
  4306. - Round a few static string arrays up to 4 byte boundaries for ARM.
  4307. - Display best share diff for scrypt as well.
  4308. - Show the best diff share as "best share" and add info to the README.
  4309. - Display the best diff share submitted so far.
  4310. - Redundant check.
  4311. - The work struct pointer in struct pc_data in findnonce is never freed yet
  4312. there is no need to allocate it separately so make struct work a static part of
  4313. the struct pc_data. s
  4314. Version 2.8.5 - October 23, 2012
  4315. - Handle crash exceptions by trying to restart cgminer unless the --no-restart
  4316. option is used.
  4317. - Switch queued count when choosing a different pool from a failed stratum pool
  4318. in getwork thread.
  4319. - Put a mandatory 5s wait between reattempting a getwork on failure to avoid
  4320. hammering requests.
  4321. - The ATI stream / AMD APP SDK environment variables appear to only interfere
  4322. with win32 builds so bypass them.
  4323. - Make sure to check pool stratum curl exists under lock before attempting any
  4324. recv to not risk dereferencing upon attempting to reinitiate stratum.
  4325. - Avoid redefining macros and align to 4 byte boundaries.
  4326. - API - add Stratum information to pools
  4327. - update FPGA-README for MMQ
  4328. Version 2.8.4 - October 18, 2012
  4329. - Time for dynamic is in microseconds, not ms.
  4330. - x86_64 builds of mingw32 are not supported directly and should just configure
  4331. as generic mingw32 builds since they're NOT 64 bit.
  4332. - Cope with both ATI stream and AMD APP SDK roots being set when building.
  4333. - Use 3 significant digits when suffix string is used and values are >1000.
  4334. - MMQ new initialisation (that works) and clocking control
  4335. - Get rid of unused warning for !scrypt.
  4336. - Use select on stratum send to make sure the socket is writeable.
  4337. - Cope with dval being zero in suffix_string and display a single decimal place
  4338. when significant digits is not specified but the value is greater than 1000.
  4339. - Pad out the suffix string function with zeroes on the right.
  4340. - Failure to calloc in bin2hex is a fatal failure always so just check for that
  4341. failure within the function and abort, simplifying the rest of the code.
  4342. - Provide locking around the change of the stratum curl structures to avoid
  4343. possible races.
  4344. - Bump opencl kernel version numbers.
  4345. - Remove atomic ops from opencl kernels given rarity of more than once nonce on
  4346. the same wavefront and the potential increased ramspeed requirements to use the
  4347. atomics.
  4348. - Clear the pool idle flag in stratum when it comes back to life.
  4349. - Display correct share hash and share difficulty with scrypt mining.
  4350. - Use explicit host to BE functions in scrypt code instead of hard coding
  4351. byteswap everywhere.
  4352. - Show work target diff for scrypt mining.
  4353. - Ease the checking on allocation of padbuffer8 in the hope it works partially
  4354. anyway on an apparently failed call.
  4355. - Watch for buffer overflows on receiving data into the socket buffer.
  4356. - Round target difficulties down to be in keeping with the rounding of detected
  4357. share difficulties.
  4358. - Dramatically simplify the dynamic intensity calculation by oversampling many
  4359. runs through the opencl kernel till we're likely well within the timer
  4360. resolution on windows.
  4361. - String alignment to 4 byte boundaries and optimisations for bin<->hex
  4362. conversions.
  4363. - In opencl_free_work, make sure to still flush results in dynamic mode.
  4364. - Align static arrays to 4 byte boundaries to appease ARM builds for stratum.
  4365. Version 2.8.3 - October 12, 2012
  4366. - Left align values that are suffix_string generated.
  4367. - Share_diff should not be converting the work data to hex.
  4368. - Off by one error.
  4369. - Prevent overflows of the port char array in extract_sockaddr.
  4370. - Disable stratum detection with scrypt.
  4371. - Use the suffix string function when displaying device hashrates.
  4372. - Be consistent with the get_statline function.
  4373. - Use the suffix string function for displaying hashrate with 4 significant
  4374. digits.
  4375. - Display the actual share diff next to the pool required diff, using a suffix
  4376. creation function to prevent values of >1000 being shown in their entirety.
  4377. - Fix 4 * 0 being 0 that would break dynamic intensity mode.
  4378. - Fix wrong byteswap macro being used on mingw32 which was breaking target
  4379. generation on stratum.
  4380. Version 2.8.2 - October 11, 2012
  4381. - Reinstate the history on dynamic intensity mode to damp fluctuations in
  4382. intensity but use an upper limit on how much the value can increase at any time
  4383. to cope with rare overflows.
  4384. - Create a fix-protocol option which prevents cgminer from switching to stratum
  4385. if it's detected.
  4386. - Simplify target generation code.
  4387. - Add support for client.get_version for stratum.
  4388. - Use a 64 bit unsigned integer on the diff target to generate the hex target.
  4389. - Update reconnect message to show whole address including port.
  4390. - Look for null values and parse correct separate array entries for url and port
  4391. with client reconnect commands for stratum.
  4392. - The command for stratum is client.reconnect, not mining.reconnect.
  4393. - Only copy the stratum url to the rpc url if an rpc url does not exist.
  4394. - Implement rudimentary mining.reconnect support for stratum.
  4395. - Ignore the value of stratum_active on calling initiate_stratum and assume
  4396. we're always trying to reinitiate it, and set the active flag to false in that
  4397. function.
  4398. - stratum auth can be unset if we fail to authorise on subsequent calls to
  4399. auth_stratum which undoes the requirement of setting it in one place so set it
  4400. in pool_active.
  4401. Version 2.8.1 - October 8, 2012
  4402. - Use the stratum url as the rpc url advertised if we switch to it.
  4403. - Count an invalid nonce count as a hardware error on opencl.
  4404. - Count each stratum work item as local work.
  4405. - Cope with one stratum pool being the only active pool when it dies by sleeping
  4406. for 5 seconds before retrying to get work from it instead of getting work
  4407. indefinitely.
  4408. - Detect stratum outage based on either select timing out or receiving an empty
  4409. buffer and properly re-establish connection by disabling the stratum_active
  4410. flag, coping with empty buffers in parse_stratum.
  4411. Version 2.8.0 - October 7, 2012
  4412. - Major upgrade - support for the stratum mining protocol.
  4413. - Fix various modminer warnings on mingw.
  4414. - Fix sign warning on windows build for bitforce.
  4415. - Cast socketfail to integer since SOCKET is an unsigned int on windows.
  4416. - Use strtod not strtol for bitforce temp backup.
  4417. - Cope with broken drivers returning nonsense values for bitforce temperatures.
  4418. - Minor warning fixes.
  4419. - Use the stratum thread to detect when a stratum pool has died based on no
  4420. message for 2 minutes.
  4421. - Only set the stratum auth flag once and once the stratum thread is started,
  4422. use that to set/unset the stratum active flag.
  4423. - Only hand off to stratum from getwork if we succeed in initiating the
  4424. protocol.
  4425. - Target should only be 32 bytes copied.
  4426. - Use a static array for work submission data instead of stack memory.
  4427. - Clear the buffer data before sprinting to it.
  4428. - Clear work stratum strings before setting them and add them to debug output.
  4429. - Drop stratum connect failed message to verbose level only since it's a regular
  4430. probing message.
  4431. - TCP Keepalive in curl is only in very recent versions and not required with
  4432. regular messages on stratum anyway.
  4433. - Move stratum sockets to curl infrastructure with locking around send+recv to
  4434. begin support for proxies and ssl.
  4435. - Make detect stratum fail if a proxy has been set up.
  4436. - Stratum does not currently have any proxy support so do not try to switch to
  4437. stratum if a proxy has been specified.
  4438. - Windows doesn't work with MSG_PEEK on recv so move to a continuously updating
  4439. buffer for incoming messages.
  4440. - Alloca is unreliable on windows so use static arrays in util.c stratum code.
  4441. - Begin support for mingw stratum build.
  4442. - Add space to reject reason.
  4443. - Parse the reject reason where possible from stratum share submission.
  4444. - Pass json error value to share result function to be able to parse reject
  4445. reason in stratum.
  4446. - Don't try to parse unneeded parameters in response to mining.subscribe.
  4447. - Remove the sshare hash entry if we failed to send it.
  4448. - Change notify message to info level to avoid spamming repeatedly when a pool
  4449. is down.
  4450. - Check the stratum pool difference has not changed compared to the work diff
  4451. when testing whether a share meets the target or not and retarget if necessary.
  4452. - Bit error in target calculation for stratum.
  4453. - Set work_block in gen_stratum_work for when work is reused to avoid thinking
  4454. it's all stale.
  4455. - Offset the current block detection to the prev block hash.
  4456. - We should be testing for id_val, not id in parse stratum response.
  4457. - Make target on stratum scale to any size by clearing sequential bits according
  4458. to diff.
  4459. - Correct target calculation in gen_stratum_work.
  4460. - If a share result has an error code but still has an id, it is likely a
  4461. reject, not an error.
  4462. - Initiate stratum the first time in pool_active only, allowing us to switch to
  4463. it on getting a failed getwork and detecting the presence of stratum on the url
  4464. at that time.
  4465. - Use 5 second timeout on sock full for now as a temporary workaround.
  4466. - If no stratum url is set by the end of the detect stratum routine, copy the
  4467. sockaddr url.
  4468. - Make all buffers slightly larger to prevent overflow.
  4469. - Make the stratum recv buffer larger than the recvsize.
  4470. - Userpass needs to be copied to user and pass earlier to allow stratum
  4471. authorisation to work with it.
  4472. - Store a sockaddr url of the stripped url used in determining sockaddr to not
  4473. confuse it with the stratum url and fix build warnings.
  4474. - Decrease the queued count with stratum work once it's staged as well.
  4475. - Allow the stratum retry to initiate and auth stratum in pool_alive to make
  4476. sure the stratum thread is started.
  4477. - Avoid duplicating pool->rpc_url and setting pool->stratum_url twice to itself.
  4478. - Detect if a getwork based pool has the X-Stratum header on startup, and if so,
  4479. switch to the stratum based pool.
  4480. - Comment update.
  4481. - Minor message change.
  4482. - Create a work item from a "clean" request from stratum allowing the new block
  4483. to be detected and the appropriate block change message to be given.
  4484. - Use statically allocated stratum strings in struct work to cope with the
  4485. inability to safely deallocate dynamically allocated ram.
  4486. - Use the current pool when deciding whether to reuse work from a stratum source
  4487. rather than the work's previous pool.
  4488. - Copy the stratum url to the rpc url to avoid none being set.
  4489. - Provide locking around stratum send operations to avoid races.
  4490. - Submit shares from stratum through the abstracted submit share function
  4491. detecting what message they belong to and showing the data from the associated
  4492. work, and then deleting it from the hash.
  4493. - Use a more robust mechanism to obtain a \n terminated string over a socket.
  4494. - Abstract out share submit as a function to be useable by stratum.
  4495. - Rename parse_stratum to parse_method as it is only for stratum messages that
  4496. contain methods.
  4497. - Display stratum as mechanism in status line when current pool is running it.
  4498. - Count each stratum notify as a getwork equivalent.
  4499. - Correct nonce submitted with share.
  4500. - Extranonce2 should be added before coinbase2.
  4501. - We should be hashing the binary coinbase, not the hex one.
  4502. - Fix endianness of nonce submitted for stratum.
  4503. - Check that stratum is already active in initiate_stratum to avoid
  4504. de-authorising ourselves by subscribing again.
  4505. - Begin implementing a hash database of submissions and attempt sending results.
  4506. - Copy parameters from stratum work required for share submission.
  4507. - Set lagging flag on first adding a pool to prevent pool slow warning at
  4508. startup.
  4509. - Fix work->target being a 32 byte binary in gen_stratum_work.
  4510. - Store and display stripped url in its own variable.
  4511. - Create machinery to divert work requests to stratum.
  4512. - Generate the work target in gen_stratum_work, setting default diff to 1 in
  4513. case it is not yet set.
  4514. - Generate work data, midstate and hash1 in gen_stratum_work.
  4515. - Generate header created from stratum structures in gen_stratum_work.
  4516. - Generate merkle root hash in gen_stratum_work.
  4517. - Generate the coinbase for generation of stratum based work.
  4518. - The number of transactions is variable so make merkle a variable length
  4519. dynamically allocated array and track how many there are for stratum.
  4520. - Rename nonce2 to n2size reflecting that it's a size variable and not the
  4521. actual nonce.
  4522. - Provide rudimentary support for stratum clean work command in the stratum
  4523. thread.
  4524. - Cope with pools being removed in the stratum thread.
  4525. - Use the pool sock value directly in the stratum thread in case it changes
  4526. after reconnecting.
  4527. - Create a stratum thread per pool that has stratum that monitors the socket and
  4528. serves received data.
  4529. - Check return value of stratum_parse.
  4530. - Complete authorisation in stratum.
  4531. - Implement stratum parsing of notify parameters and storing them in the pool
  4532. stratum work structure.
  4533. - Create helper functions for duplicating json strings to avoid keeping json
  4534. references in use.
  4535. - Append \n in the sock_send function instead of adding it when constructing
  4536. json in stratum.
  4537. - Don't keep any json references around with stratum structures.
  4538. - Create parse_stratum function that hands off stratum parameters to other
  4539. functions to manage pool stratum work struct variables. Implement mining
  4540. difficulty setting.
  4541. - Create helper functions for checking when a socket is ready to read on and
  4542. receive a single line at a time. Begin stratum authorisation process.
  4543. - Provide a helper function for reading a single \n terminated string from a
  4544. socket.
  4545. - Create a stratum work structure to store current work variables.
  4546. - Test specifically for stratum being active in pool_active.
  4547. - Detect stratum in common place when adding urls, and use a bool to tell us
  4548. when it's active.
  4549. - Fix warnings.
  4550. - Extract and store various parameters on stratum init confirming successful
  4551. mining notify.
  4552. - Use existing socket macros and close the socket on failure in init stratum.
  4553. - Initiate stratum and grab first json result.
  4554. - Get detailed addressinfo from the parsed URL for future raw socket usage when
  4555. possible. IPV4 only for now.
  4556. - Prepare for getaddrinfo call.
  4557. - Add data structures to pool struct for socket communications.
  4558. - Put all socket definitions in util.h to allow reusing by added socket
  4559. functions to be used in util.c.
  4560. Version 2.7.7 - October 7, 2012
  4561. - Fix unused warnings on ming build.
  4562. - Fix sign warning in ocl.c
  4563. - fds need to be zeroed before set in modminer.
  4564. - Put scrypt warning on separate line to avoid 0 being shown on windows as
  4565. bufsize.
  4566. - Display correct pool number when block is found.
  4567. - Prevent corrupt values returned from the opencl code from trying to read
  4568. beyond the end of the buffer by masking the value to a max of 15.
  4569. - Icarus USB write failure is also a comms error
  4570. - api.c DEBUG message has no paramter
  4571. - Icarus catch more USB errors and close/reopen the port
  4572. - API-README update cgminer verison number
  4573. - hashmeter fix stats kh/s on 32bit windows
  4574. Version 2.7.6 - September 24, 2012
  4575. - Reorder libztex header include order to fix missing struct definition.
  4576. - Display share difficulty on log with a shortened hash display on submission.
  4577. - API stats add some pool getwork difficulty stats
  4578. - Ignore any pings pushed to the worker threads if the thread is still paused to
  4579. prevent it being enabled and disabled repeatedly.
  4580. - README - FAQ - usermod group - shouldn't remove other groups
  4581. - Test for sequential getwork failures on a pool that might actually be up but
  4582. failing to deliver work as we may end up hammering it repeatedly by mistake.
  4583. - reduce windows compile warnings
  4584. - util.c - bug - proxy - no data end condition
  4585. - As we average gpu time over 5 work intervals for dynamic GPU intensity, there
  4586. is no need to maintain a rolling average and it avoids the potential long term
  4587. corruption of a single overflow value.
  4588. - Test for the now-automatically exported variable AMDAPPSDKROOT when looking
  4589. for the presence of the OpenCL headers.
  4590. - API don't change 'Diff1 Shares' - backward compatability FTW
  4591. - miner.php highlighting correctly handling difficulty
  4592. - API - Add last share difficulty for devices and pool
  4593. - Store and report Accepted,Rejected,Stale difficulty in the summary and API
  4594. - WorkTime - display prevblock for scrypt
  4595. - api.c remove compile warnings
  4596. - Calculate work difficulty for each getwork and display with WorkTime debug
  4597. - remove MMQ unused variable warning
  4598. - FPGA - allow long or short device names in detect code + style police
  4599. - WorkTime - multiple nonce per work and identify the work source
  4600. - Optional WorkTime details with each Accepted/Rejected work item
  4601. - Icarus - ignore hardware errors in timing mode
  4602. - miner.php oops - mistype
  4603. - miner.php by default don't display IP/Port numbers in error messages
  4604. - api.c all STATUS messages automatically escaped
  4605. - api.c add missing escape for comma in MSG_PGAUNW
  4606. - API add display of and setting queue,scantime,expiry
  4607. - HW: dont submit bad shares
  4608. - save individual pool proxy settings to config
  4609. - --default-config - allow command line to define the default configuration file
  4610. for loading and saving
  4611. - API-README update for pools proxy info
  4612. - README URL proxy must use quote so show in the example
  4613. - bug: remove proxy: from the front of the proxy used
  4614. - CURL support for individual proxy per pool and all proxy types
  4615. - README spelling/etc
  4616. - README - FPGA device FAQ
  4617. - HW: error counter auto for all devices - ztex code not fixed
  4618. - API pgaidentify - unsupported message should be a warning
  4619. - API/BFL identify a device - currently only BFL to flash the led
  4620. - BFL add throttle count to internal stats + API
  4621. - BFL: missing device id in log message
  4622. - miner.php correct to new Diff1 Work field names
  4623. - API add device diff1 work
  4624. - API-README update
  4625. - api.c Correct diff1 field name
  4626. - count device diff1 shares
  4627. - API-README more debug parameter information
  4628. - API allow full debug settings control
  4629. Version 2.7.5 - August 31, 2012
  4630. - Adjust opencl intensity when adjusting thread count to prevent it getting
  4631. pegged at a value below the minimum threads possible.
  4632. - miner.h max_hashes -> int64_t
  4633. - Keep the local block number in the blocks structs stored and sort them by
  4634. number to guarantee we delete the oldest when ageing the block struct entries.
  4635. - Use correct sdk version detection for SDK 2.7
  4636. - Revert "Pick worksize 256 with Cypress if none is specified."
  4637. - Test for lagging once more in queue_request to enable work to leak to backup
  4638. pools.
  4639. - There is no need to try to switch pools in select_pool since the current pool
  4640. is actually not affected by the choice of pool to get work from.
  4641. - Only clear the pool lagging flag if we're staging work faster than we're using
  4642. it.
  4643. - needed flag is currently always false in queue_request. Remove it for now.
  4644. - thr is always NULL going into queue_request now.
  4645. Version 2.7.4 - August 23, 2012
  4646. - Perform select_pool even when not lagging to allow it to switch back if needed
  4647. to the primary.
  4648. - Simplify macros in output kernels avoiding apparent loops and local variables.
  4649. - Carry the needed bool over the work command queue.
  4650. - Move the decision to queue further work upstream before threads are spawned
  4651. based on fine grained per-pool stats and increment the queued count immediately.
  4652. - Track queued and staged per pool once again for future use.
  4653. - OpenCL 1.0 does not have native atomic_add and extremely slow support with
  4654. atom_add so detect opencl1.0 and use a non-atomic workaround.
  4655. - Pools: add RollTime info to API 'stats' and 'Stats' button in miner.php
  4656. Version 2.7.3 - August 22, 2012
  4657. - Minimise the number of getwork threads we generate.
  4658. Version 2.7.2 - August 22, 2012
  4659. - Pick worksize 256 with Cypress if none is specified.
  4660. - Give warning with sdk2.7 and phatk as well.
  4661. - Whitelist sdk2.7 for diablo kernel as well.
  4662. - Only keep the last 6 blocks in the uthash database to keep memory usage
  4663. constant. Storing more is unhelpful anyway.
  4664. - BFL Flash - always distribute source
  4665. - Increase kernel versions signifying changed APIs.
  4666. - BFL flash - include source in builds and more FPGA-README
  4667. - Check we haven't staged work while waiting for a curl entry before proceeding.
  4668. - Use atomic ops to never miss a nonce on opencl kernels, including nonce==0,
  4669. also allowing us to make the output buffer smaller.
  4670. - Remove compile errors/warnings and document compile/usage in FPGA-README
  4671. - bitforce-firmware-flash.c by Luke-jr
  4672. - Ignore the submit_fail flag when deciding whether to recruit more curls or not
  4673. since we have upper bounds on how many curls can be recruited, this test is
  4674. redundant and can lead to problems.
  4675. - API-README update cgminer version number
  4676. - API-README fix groups P: example mistake
  4677. - API-README add COIN and other edits
  4678. - gpu->hit should be reset on new work as well.
  4679. - Do not add time to dynamic opencl calculations over a getwork.
  4680. - miner.php allow 'coin' is custom pages
  4681. Version 2.7.1 - August 21, 2012
  4682. - Update windows build instructions courtesy of sharky.
  4683. - Increase max curls to number of mining threads + queue * 2, accounting for up
  4684. and downstream comms.
  4685. - Queue enough requests to get started.
  4686. - There is no point trying to clone_work in get_work() any more since we clone
  4687. on every get_work_thread where possible.
  4688. - There is no point subtracting 1 from maxq in get_work_thread.
  4689. - Only set lagging flag once there are no staged work items.
  4690. - select_pool does not switch back to the primary once lagging is disabled.
  4691. - miner.php allow page title to be defined in myminer.php
  4692. - Free work before retrying in get_work_thread.
  4693. - Increment total work counter under mutex lock.
  4694. - Increment the queued count after the curl is popped in case there's a delay
  4695. waiting on curls and we think we've queued work when in fact we're waiting
  4696. - API new command 'coin' with mining information
  4697. - Do the dynamic timing in opencl code over a single pass through scanhash to
  4698. make sure we're only getting opencl times contributing to the measured inte
  4699. - Increase curl reaping time to 5 minutes since comms between curl requests can
  4700. be 2 mins apart with lots of rolltime.
  4701. - No need for extra variable in hash_push.
  4702. - Remove short options -r and -R to allow them to be reused and remove readme
  4703. entries for deprecated options.
  4704. - Avoid attempting to recursively lock the console mutex by disabling warnings
  4705. in gpu_fanpercent when fanspeed monitoring fails on windows. Debugged by l
  4706. - Deprecate the opt_fail_pause parameter, leaving a null placeholder for
  4707. existing configurations.
  4708. - Don't pause after failed getwork, set lagging flag and reassess.
  4709. - Add message to share if it's a resubmit.
  4710. - We should not be pausing in trying to resubmit shares.
  4711. - Get rid of the extending fail pause on failed connects since we discard work
  4712. after a period.
  4713. - get_work always returns true so turn it into a void function.
  4714. - get_work never returns false so get rid of fail pause loop.
  4715. - Get rid of pause and retry from get_upstream_work so we only do it from one
  4716. place.
  4717. - Deprecate the opt_retries feature as no one wants cgminer to automatically
  4718. abort. Leave a null placeholder for configurations that still have it.
  4719. - Reinstate fix ADL gpu-map not working when there are more ADL devices than
  4720. openCL patch by Nite69. Add virtual adl mapping for when none is specified o
  4721. - miner.php show summary Diff1 Shares total
  4722. - miner.php fix Work Utility totals
  4723. - miner.php format new Work Utility and Diff1 Shares
  4724. - API V1.17 show Work Utility and Diff1 Shares
  4725. Version 2.7.0 - August 18, 2012
  4726. - Introduce a new statistic, Work Utility, which is the number of difficulty 1
  4727. shares solved per minute. This is useful for measuring a relative rate of work
  4728. that is independent of reject rate and target difficulty.
  4729. - Implement a new pool strategy, BALANCE, which monitors work performed per pool
  4730. as a rolling average every 10 minutes to try and distribute work evenly over all
  4731. the pools. Do this by monitoring diff1 solutions to allow different difficulty
  4732. target pools to be treated equally, along with solo mining. Update the
  4733. documentation to describe this strategy and more accurately describe the
  4734. load-balance one.
  4735. - Getwork fail was not being detected. Remove a vast amount of unused variables
  4736. and functions used in the old queue request mechanism and redefine the getfail
  4737. testing.
  4738. - Don't try to start devices that don't support scrypt when scrypt mining.
  4739. - 0 is a valid return value for read so only break out if read returns -1.
  4740. - Consider us lagging only once our queue is almost full and no staged work.
  4741. - Simplify the enough work algorithm dramatically.
  4742. - Only queue from backup pools once we have nothing staged.
  4743. - Don't keep queueing work indefinitely if we're in opt failover mode.
  4744. - Make sure we don't opt out of queueing more work if all the queued work is
  4745. from one pool.
  4746. - Set lagging flag if we're on the last of our staged items.
  4747. - Reinstate clone on grabbing work.
  4748. - Grab clones from hashlist wherever possible first.
  4749. - Cull all the early queue requests since we request every time work is popped
  4750. now.
  4751. - Keep track of staged rollable work item counts to speed up clone_available.
  4752. - Make expiry on should_roll to 2/3 time instead of share duration since some
  4753. hardware will have very fast share times.
  4754. - Do the cheaper comparison first.
  4755. - Check that we'll get 1 shares' worth of work time by rolling before saying we
  4756. should roll the work.
  4757. - Simplify all those total_secs usages by initialising it to 1 second.
  4758. - Overlap queued decrementing with staged incrementing.
  4759. - Artificially set the pool lagging flag on pool switch in failover only mode as
  4760. well.
  4761. - Artificially set the pool lagging flag on work restart to avoid messages about
  4762. slow pools after every longpoll.
  4763. - Factor in opt_queue value into enough work queued or staged.
  4764. - Roll work whenever we can on getwork.
  4765. - Queue requests for getwork regardless and test whether we should send for a
  4766. getwork from the getwork thread itself.
  4767. - Get rid of age_work().
  4768. - 0 is a valid return value for read so only break out if read returns -1.
  4769. - Offset libusb reads/writes by length written as well in ztex.
  4770. - Cope with timeouts and partial reads in ztex code.
  4771. - fpga serial I/O extra debug (disabled by default)
  4772. Version 2.6.5 - August 15, 2012
  4773. - Don't try to get bitforce temperature if we're polling for a result to
  4774. minimise the chance of interleaved responses.
  4775. - Set memory clock based on memdiff if present from with engine changes,
  4776. allowing it to parallel manual changes from the menu as well.
  4777. - Increase the timeout on bitforce as per Paul Sheppard's suggestion to account
  4778. for throttling + work time + excess.
  4779. - Fix ADL gpu-map not working when there are more ADL devices than openCL.
  4780. Initial patch supplied by Nite69. Modified to suit.
  4781. - Windows' timer resolution is limited to 15ms accuracy. This was breaking
  4782. dynamic intensity since it tries to measure below this. Since we are repeatedly
  4783. sampling similar timeframes, we can average the gpu_us result over 5 different
  4784. values to get very fine precision.
  4785. - Fix harmless unused warnings in scrypt.h.
  4786. - api.c typo
  4787. - API allow display/change failover-only setting
  4788. - Check we are not lagging as well as there is enough work in getwork.
  4789. - Minimise locking and unlocking when getting counts by reusing shared mutex
  4790. lock functions.
  4791. - Avoid getting more work if by the time the getwork thread is spawned we find
  4792. ourselves with enough work.
  4793. - The bitforce buffer is cleared and hw error count incremented on return from a
  4794. failed send_work already so no need to do it within the send_work function.
  4795. - miner.php allow a custom page section to select all fields with '*' - e.g. to
  4796. create a STATS section on a custom page
  4797. - Escape " and \ when writing json config file
  4798. - miner.php optional single rig totals (on by default)
  4799. Version 2.6.4 - August 7, 2012
  4800. - Convert the serial autodetect functions to use int instead of char to
  4801. enumerate devices.
  4802. - Make the serial open timeout for BFL generically 1 second on windows.
  4803. - Deuglify windows autodetect code for BFL.
  4804. - There is no point zeroing temperature in BFL if we fail to get a response, and
  4805. we should register it as a HW error, suggesting throttling.
  4806. - Update SCRYPT README with information about HW errors.
  4807. - Use the scrypt CPU code to confirm results from OCL code, and mark failures as
  4808. HW errors, making it easier to tune scrypt parameters.
  4809. - We may as well leave one curl still available per pool instead of reaping the
  4810. last one.
  4811. - Need to recheck the pool->curls count on regaining the pool lock after the
  4812. pthread conditional wait returns.
  4813. - Display reaped debug message outside mutex lock to avoid recursive locking.
  4814. - Add specific information when ADL detects error -10 saying the device is not
  4815. enabled.
  4816. - api.c update API start message and include port number
  4817. - miner.php ignore arg when readonly
  4818. - miner.php allow pool inputs: delete, addpool, poolpriority
  4819. Version 2.6.3 - August 5, 2012
  4820. - Count likely throttling episodes on bitforce devices as hardware errors.
  4821. - Style cleanups.
  4822. - Use FTD2XX.DLL on Windows to autodetect BitFORCE SHA256 devices.
  4823. - Make pool_disabled the first in the enums == 0, fixing the pool enabled count
  4824. which compares if value is not enabled before enabling it.
  4825. - Correct writing of scrypt parameters to config file based on command line
  4826. parameters only.
  4827. - Use different variables for command line specified lookup gap and thread
  4828. concurrency to differentiate user defined versus auto chosen values.
  4829. - Queue a request on pool switch in case we have no work from the new pool yet.
  4830. - Display failover only mode in pool menu and allow it to be toggled live.
  4831. - Reinstate check for system queueing lag when the current pool's queue is maxed
  4832. out, there is no staged work, and the work is needed now.
  4833. - There is no need for pool active testing to be mandatory any more with queue
  4834. request changes.
  4835. - Fix harmless warnings.
  4836. - Check the current staged and global queued as well before queueing requests.
  4837. Discard stales before ageing work in the watchdog thread. Queue requests after
  4838. discarding and ageing work in watchdog thread. Display accurate global queued in
  4839. curses output. Reuse variable in age_work().
  4840. - The queueing mechanism has become a complex state machine that is no longer
  4841. predictable. Rewrite it from scratch watching only current queues in flight and
  4842. staged work available on a pool by pool basis.
  4843. - API remove unused warning in non-GPU compile
  4844. - api.c in linux allow to open a closed socket in TIME_WAIT
  4845. - Queue an extra request whenever staged work drops below mining thread count in
  4846. hash_pop.
  4847. - Update debian package configs to v2.6.2
  4848. Version 2.6.2 - August 3, 2012
  4849. - Scrypt mining does not support block testing yet so don't try to print it.
  4850. - Clear the bitforce buffer whenever we get an unexpected result as it has
  4851. likely throttled and we are getting cached responses out of order, and use the
  4852. temperature monitoring as a kind of watchdog to flush unexpected results.
  4853. - It is not critical getting the temperature response in bitforce so don't
  4854. mandatorily wait on the mutex lock.
  4855. - Check there is a cutoff temp actually set in bitforce before using it as a cut
  4856. off value otherwise it may think it's set to zero degrees.
  4857. - We dropped the temporary stopping of curl recruiting on submit_fail by
  4858. mistake, reinstate it.
  4859. - Make threads report in either side of the scanhash function in case we miss
  4860. reporting in when restarting work.
  4861. - Don't make mandatory work and its clones last forever.
  4862. - Make test work for pool_active mandatory work items to smooth out staged work
  4863. counts when in failover-only mode.
  4864. - Add debugging output when work is found stale as to why.
  4865. - Print the 3 parameters that are passed to applog for a debug line in
  4866. bitforce.c
  4867. - Clear bitforce buffer on init as previously.
  4868. - Add some headroom to the number of curls available per pool to allow for
  4869. longpoll and sendwork curls.
  4870. - Revert "Revert "Change BFL driver thread initialising to a constant 100ms
  4871. delay between devices instead of a random arrangement.""
  4872. - Revert "Remove bitforce_thread_init"
  4873. - Show the correct base units on GPU summary.
  4874. - Differentiate between the send return value being a bool and the get return
  4875. value when managing them in bitforce scanhash.
  4876. - 23a8c60 Revert "bitforce: Skip out of sending work if work restart requested"
  4877. Version 2.6.1 - July 30, 2012
  4878. - Display scrypt as being built in as well.
  4879. - Fix build warning about KL_SCRYPT when built without scrypt support.
  4880. - Remove the low hash count determinant of hardware being sick. A low hash rate
  4881. can be for poor network connectivity or scrypt mining, neither of which are due
  4882. to a sick device.
  4883. - api.c poolpriority changes
  4884. Version 2.6.0 - July 29, 2012
  4885. - Display kilohash when suitable, but store the global mhash value still truly
  4886. in megahashes to not break the API output.
  4887. - Don't try and print curses output for devices that won't fit on the screen.
  4888. - Add scrypt documentation in the form of a separate readme.
  4889. - Fix build error without scrypt enabled.
  4890. - Limit total number of curls recruited per pool to the number of mining threads
  4891. to prevent blasting the network when we only have one pool to talk to.
  4892. - bitforce: Skip out of sending work if work restart requested
  4893. - Keep a counter of enabled pools and use that instead of iterating over the
  4894. pool list. Use that value to ensure we don't set the last remaining active pool
  4895. to the rejecting state.
  4896. - fpgautils: add support for 57.6 kBd serial
  4897. - miner.php add a socket RCV timeout for if cgminer is hung and the API thread
  4898. is still running
  4899. - Limit thread concurrency for scrypt to 5xshaders if shaders is specified.
  4900. - Simplify repeated use of gpus[gpu]. in ocl.c
  4901. - Find the nearest power of 2 maximum alloc size for the scrypt buffer that can
  4902. successfully be allocated and is large enough to accomodate the thread
  4903. concurrency chosen, thus mapping it to an intensity.
  4904. - Don't make opt_scrypt mandatory blocking with opencl code.
  4905. - Update kernel versions reflecting changes in the API.
  4906. - Make the thread concurrency and lookup gap options hidden on the command line
  4907. and autotune parameters with a newly parsed --shaders option.
  4908. - Fix target testing with scrypt kernel as it would have been missing shares
  4909. below target.
  4910. - Bugfix: Use a mutex to control non-curses output
  4911. - Simplify code to a single vprintf path for curses-less printing
  4912. - Move opt_quiet check to my_log_curses, so it works for curses-less builds
  4913. - Use log_generic for vapplog to cut down on code duplication
  4914. - Add space to log output now that there is more screen real estate available.
  4915. - BFL force all code to timeout to avoid hanging
  4916. - Bugfix: Copy argv[0] given to dirname()
  4917. - Always create the largest possible padbuffer for scrypt kernels even if not
  4918. needed for thread_concurrency, giving us some headroom for intensity levels.
  4919. - Use the detected maximum allocable memory on a GPU to determine the optimal
  4920. scrypt settings when lookup_gap and thread_concurrency parameters are not given.
  4921. - Check the maximum allocable memory size per opencl device.
  4922. - Add debugging output if buffer allocation fails for scrypt and round up
  4923. bufsize to a multiple of 256.
  4924. - Nonce testing for btc got screwed up, leading to no accepted shares. Fix it.
  4925. - Display size of scrypt buffer used in debug.
  4926. - Allow intensities up to 20 if scrypt is compiled in.
  4927. - Add name to scrypt kernel copyright.
  4928. - Allow lookup gap and thread concurrency to be passed per device and store
  4929. details in kernel binary filename.
  4930. - Ignore negative intensities for scrypt.
  4931. - Change the scale of intensity for scrypt kernel and fix a build warning.
  4932. - Correct target value passed to scrypt kernel.
  4933. - Use 256 output slots for kernels to allow 1 for each worksize.
  4934. - Test the target in the actual scrypt kernel itself saving further
  4935. calculations.
  4936. - Reinstate GPU only opencl device detection.
  4937. - Decrease lookup gap to 1. Does not seem to help in any way being 2.
  4938. - Fix build.
  4939. - Make pad0 and pad1 local variable in scrypt kernel.
  4940. - Constify input variable in scrypt kernel.
  4941. - Send correct values to scrypt kernel to get it finally working.
  4942. - Create command queue before compiling program in opencl.
  4943. - Detach pthread from within the api thread in case it is terminated due to not
  4944. being instantiated before pthread_cancel is called from main, leading to a
  4945. segfault.
  4946. - Debug output per thread hashrate is out by a factor of 1000.
  4947. - Initialise mdplatform.
  4948. - Find the gpu platform with the most devices and use that if no platform option
  4949. is passed.
  4950. - Allow more platforms to be probed if first does not return GPUs.
  4951. - Fix external scrypt algo missing.
  4952. - Limit scrypt to 1 vector.
  4953. - Handle KL_SCRYPT in config write.
  4954. - Get rid of stuff.
  4955. - Don't enqueuewrite buffer at all for pad8 and pass work details around for
  4956. scrypt in dev_blk.
  4957. - Set the correct data for cldata and prepare for pad8 fixes.
  4958. - Bugfix: Fix build without curses but with OpenCL
  4959. - Find the gpu platform with the most devices and use that if no platform option
  4960. is passed.
  4961. - Allow more platforms to be probed if first does not return GPUs.
  4962. - Get rid of spaces in arrays in scrypt kernel.
  4963. - Start with smaller amount of hashes in cpu mining to enable scrypt to return
  4964. today sometime.
  4965. - Show Khash hashrates when scrypt is in use.
  4966. - Free the scratchbuf memory allocated in scrypt and don't check if CPUs are
  4967. sick since they can't be. Prepare for khash hash rates in display.
  4968. - Add cpumining capability for scrypt.
  4969. - Set scrypt settings and buffer size in ocl.c code to be future modifiable.
  4970. - Cope with when we cannot set intensity low enough to meet dynamic interval by
  4971. inducing a forced sleep.
  4972. - Make dynamic and scrypt opencl calls blocking.
  4973. - Calculate midstate in separate function and remove likely/unlikely macros
  4974. since they're dependent on pools, not code design.
  4975. - bitforce: Use "full work" vs "nonce range" for kernel name
  4976. - Display in debug mode when we're making the midstate locally.
  4977. - Fix nonce submission code for scrypt.
  4978. - Make sure goffset is set for scrypt and drop padbuffer8 to something
  4979. manageable for now.
  4980. - Set up buffer8 for scrypt.
  4981. - Build fix for opt scrypt.
  4982. - Don't check postcalc nonce with sha256 in scrypt.
  4983. - Don't test nonce with sha and various fixes for scrypt.
  4984. - Make scrypt buffers and midstate compatible with cgminer.
  4985. - Use cgminer specific output array entries in scrypt kernel.
  4986. - Provide initial support for the scrypt kernel to compile with and mine scrypt
  4987. with the --scrypt option.
  4988. - Enable completely compiling scrypt out.
  4989. - Begin import of scrypt opencl kernel from reaper.
  4990. - bitforce_get_result returns -1 on error now.
  4991. - Check return value of read in BFgets
  4992. - Bugfix: Make our Windows nanosleep/sleep replacements standards-compliant
  4993. (which fixes nmsleep) and include compat.h for bitforce (for sleep)
  4994. - rpc: Use a single switch statement for both stringifications of cgpu->status
  4995. - Fix whitespace mangling.
  4996. - miner.php fix rig # when miners fail
  4997. - Only try to shut down work cleanly if we've successfully connected and started
  4998. mining.
  4999. - Use switch statement for cgpu->status and fix spelling.
  5000. - Abbrv. correction
  5001. - Bugfix: Don't declare devices SICK if they're just busy initialising
  5002. - Bugfix: Calculate nsec in nmsleep correctly
  5003. - Bugfix: Adapt OpenCL scanhash errors to driver API change (errors are now -1,
  5004. not 0)
  5005. - Remove superfluous ave_wait
  5006. - Put kname change for broken nonce-range back in
  5007. - Add average wait time to api stats
  5008. - Change BFL driver thread initialising to a constant 100ms delay between
  5009. devices instead of a random arrangement.
  5010. - Spelling typo.
  5011. - Time opencl work from start of queueing a kernel till it's flushed when
  5012. calculating dynamic intensity.
  5013. - Modify te scanhash API to use an int64_t and return -1 on error, allowing zero
  5014. to be a valid return value.
  5015. - Check for work restart after the hashmeter is invoked for we lose the hashes
  5016. otherwise contributed in the count.
  5017. - Remove disabled: label from mining thread function, using a separate
  5018. mt_disable function.
  5019. - Style changes.
  5020. - Missed one nonce-range disabling.
  5021. - Add average return time to api stats
  5022. - miner.php allow rig names in number buttons
  5023. - Remove bitforce_thread_init The delay thing does nothing useful... when long
  5024. poll comes around, all threads restart at the same time anyway.
  5025. - Change timeouts to time-vals for accuracy.
  5026. - fix API support for big endian machines
  5027. - Cope with signals interrupting the nanosleep of nmsleep.
  5028. - Use standard cfsetispeed/cfsetospeed to set baud rate on *nix
  5029. - miner.php split() flagged deprecated in PHP 5.3.0
  5030. - More BFL tweaks. Add delay between closing and reopening port. Remove buffer
  5031. clear in re-init Add kernel type (mini-rig or single)
  5032. - Make long timeout 10seconds on bitforce for when usleep or nanosleep just
  5033. can't be accurate...
  5034. Version 2.5.0 - July 6, 2012
  5035. - Fix --benchmark not working since the dynamic addition of pools and pool
  5036. stats.
  5037. - Make disabling BFL nonce range support a warning since it has to be explicitly
  5038. enabled on the command line now.
  5039. - miner.php allow renaming table headers
  5040. - Make bitforce nonce range support a command line option --bfl-range since
  5041. enabling it decrease hashrate by 1%.
  5042. - Add sanity checking to make sure we don't make sleep_ms less than 0 in
  5043. bitforce.
  5044. - The fastest minirig devices need a significantly smaller starting sleep time.
  5045. - Use a much shorter initial sleep time to account for faster devices and nonce
  5046. range working, and increase it if nonce range fails to work.
  5047. - Use nmsleep instead of usleep in bitforce.
  5048. - Provide a ms based sleep function that uses nanosleep to avoid the inaccuracy
  5049. of usleep on SMP systems.
  5050. - delay_time_ms is always set so need not be initialised in bitforce.
  5051. - Increase bitforce timeout to 10 seconds.
  5052. - Add more hysteresis and poll ~5 times to allow for timer delays in bitforce
  5053. devices.
  5054. - miner.php allow alternating line colours (off by default)
  5055. - Display the actual duration of wait when it is greater than the cutoff.
  5056. - Set nonce to maximum once we determine nonce range support is broken.
  5057. - Initial wait time is always known so no need to zero it beforehand in
  5058. bitforce.
  5059. - No point counting wait time until the work is actually sent to bitforce
  5060. devices.
  5061. - Use string comparison functions instead of explicit comparisons.
  5062. - Account for wait_ms time when nonce_range is in use on BFL.
  5063. - Split nonces up into 1/5 chunks when nonce range is supported.
  5064. - limit clear buffer iterations.
  5065. - Ad fd check to clear buffer.
  5066. - miner.php remove incorrect 'DATE' error message
  5067. - miner.php allow summary header in custom pages
  5068. - Disable nonce range support in BFL when broken support is detected.
  5069. - Restart_wait is only called with a ms value so incorporate that into the
  5070. function.
  5071. - Only try to adjust dev width when curses is built in.
  5072. - miner.php define custom sum fields as a simple array
  5073. - Fix off-by-one error in nonce increment in bfl.
  5074. - Use BE when setting nonce in bitforce nonce range work.
  5075. - Enable nonce range in the normal init sequence for bfl.
  5076. - Queue extra work at 2/3 differently depending on whether we're using nonce
  5077. range or not.
  5078. - Initially enable support for nonce range support on bfl, splitting nonces up
  5079. into 3/4 size and only disable it if it fails on work submit.
  5080. - Attempt to detect nonce range support in BFL by sending work requring its
  5081. support.
  5082. - Limit retrying on busy for up to BITFORCE_TIMEOUT_MS
  5083. - Attempt to initialise while bitforce device returns BUSY.
  5084. - Extend length of string that can be passed to BFL devices.
  5085. - Fix signedness warning.
  5086. - Adjust device width column to be consistent.
  5087. - Use cgpu-> not gpus[] in watchdog thread.
  5088. - Add api stats (sleep time)
  5089. - Timing tweaks Added long and short timeouts, short for detecting throttling,
  5090. long to give up totally. Reset sleep time when device re-initialised Still check
  5091. results after timeout Back up a larger time if result on first poll.
  5092. - Add API Notify counter 'Comms Error'
  5093. - Style police on api.c
  5094. - Do all logging outside of the bitforce mutex locking to avoid deadlocks.
  5095. - Remove applog call from bfwrite to prevent grabbing nested mutexes.
  5096. - Bitforce style changes.
  5097. - Minor style changes.
  5098. - Remove needless roundl define.
  5099. - Made JSON error message verbose.
  5100. - Fine-tune timing adjustment. Also remove old work_restart timing.
  5101. - Check for gpu return times of >= 0, not just 0, to fix intensity dropping to
  5102. -10.
  5103. - Restart is zeroed in the mining thread so no need to do it inside the bitforce
  5104. code.
  5105. - More improvements to comms. BFL return nothing when throttling, so should not
  5106. be considered an error. Instead repeat with a longer delay.
  5107. - Polling every 10ms there's not much point checking the pthread_cond_timedwait
  5108. as it just adds overhead. Simply check the value of work_restart in the bfl main
  5109. polling loop.
  5110. - Use a pthread conditional that is broadcast whenever work restarts are
  5111. required. Create a generic wait function waiting a specified time on that
  5112. conditional that returns if the condition is met or a specified time passed to
  5113. it has elapsed. Use this to do smarter polling in bitforce to abort work, queue
  5114. more work, and check for results to minimise time spent working needlessly.
  5115. - Add busy time to wait time.
  5116. - api.c put version up to 1.14
  5117. - Add tiny delay after writing to BFL Change BFL errors to something more human
  5118. readable Send work busy re-tries after 10ms delay
  5119. Version 2.4.4 - July 1, 2012
  5120. - Fix builds on non gnu platforms.
  5121. - api.c ensure old mode is always available when not using --api-groups + quit()
  5122. on param errors
  5123. - Implement rudimentary X-Mining-Hashrate support.
  5124. - Detect large swings in temperature when below the target temperature range and
  5125. change fan by amounts dependant on the value of tdiff.
  5126. - Adjust the fanspeed by the magnitude of the temperature difference when in the
  5127. optimal range.
  5128. - Revert "Restarting cgminer from within after ADL has been corrupted only leads
  5129. to a crash. Display a warning only and disable fanspeed monitoring."
  5130. - api.c fix json already closed
  5131. - implement and document API option --api-groups
  5132. - Put upper bounds to under 2 hours that work can be rolled into the future for
  5133. bitcoind will deem it invalid beyond that.
  5134. - define API option --api-groups
  5135. - api.c allow unwell devices to be enabled so they can be cured
  5136. - miner.php - fix/enable autorefresh for custom pages
  5137. - miner.php allow custom summary pages - new 'Mobile' summary
  5138. - Work around pools that advertise very low expire= time inappropriately as this
  5139. leads to many false positives for stale shares detected.
  5140. - Only show ztex board count if any exist.
  5141. - There is no need for work to be a union in struct workio_cmd
  5142. - fpgautils.c include a debug message for all unknown open errors
  5143. - Don't keep rolling work right up to the expire= cut off. Use 2/3 of the time
  5144. between the scantime and the expiry as cutoff for reusing work.
  5145. - Log a specific error when serial opens fail due to lack of user permissions
  5146. - Increase GPU timing resolution to microsecond and add sanity check to ensure
  5147. times are positive.
  5148. - Opencl code may start executing before the clfinish order is given to it so
  5149. get the start timing used for dynamic intensity from before the kernel is
  5150. queued.
  5151. - fpgautils.c - set BAUD rate according to termio spec
  5152. - fpgautils.c - linux ordering back to the correct way
  5153. - miner.php remove unneeded '.'s
  5154. - miner.php add auto refresh options
  5155. - miner.php add 'restart' next to 'quit'
  5156. - miner.php make fontname/size configurable with myminer.php
  5157. - Make the pools array a dynamically allocated array to allow unlimited pools to
  5158. be added.
  5159. - Make the devices array a dynamically allocated array of pointers to allow
  5160. unlimited devices.
  5161. - Dynamic intensity for GPUs should be calculated on a per device basis. Clean
  5162. up the code to only calculate it if required as well.
  5163. - Use a queueing bool set under control_lock to prevent multiple calls to
  5164. queue_request racing.
  5165. - Use the work clone flag to determine if we should subtract it from the total
  5166. queued variable and provide a subtract queued function to prevent looping over
  5167. locked code.
  5168. - Don't decrement staged extras count from longpoll work.
  5169. - Count longpoll's contribution to the queue.
  5170. - Increase queued count before pushing message.
  5171. - Test we have enough work queued for pools with and without rolltime
  5172. capability.
  5173. - As work is sorted by age, we can discard the oldest work at regular intervals
  5174. to keep only 1 of the newest work items per mining thread.
  5175. - Roll work again after duplicating it to prevent duplicates on return to the
  5176. clone function.
  5177. - Abstract out work cloning and clone $mining_threads copies whenever a rollable
  5178. work item is found and return a clone instead.
  5179. - api.c display Pool Av in json
  5180. - Take into account average getwork delay as a marker of pool communications
  5181. when considering work stale.
  5182. - Work out a rolling average getwork delay stored in pool_stats.
  5183. - Getwork delay in stats should include retries for each getwork call.
  5184. - Walk through the thread list instead of searching for them when disabling
  5185. threads for dynamic mode.
  5186. - Extend nrolltime to support the expiry= parameter. Do this by turning the
  5187. rolltime bool into an integer set to the expiry time. If the pool supports
  5188. rolltime but not expiry= then set the expiry time to the standard scantime.
  5189. - When disabling fanspeed monitoring on adl failure, remove any twin GPU
  5190. association. This could have been leading to hangs on machines with dual GPU
  5191. cards when ADL failed.
  5192. - modminer: Don't delay 2nd+ FPGAs during work restart
  5193. - Disable OpenCL code when not available.
  5194. - Fix openwrt crashing on regeneratehash() by making check_solve a noop.
  5195. - FPGA - allow device detect override without an open failure
  5196. - Fix sign warning.
  5197. Version 2.4.3 - June 14, 2012
  5198. - can_roll and should_roll should have no bearing on the cycle period within the
  5199. miner_thread so remove it.
  5200. - Check for strategy being changed to load balance when enabling LPs.
  5201. - Check that all threads on the device that called get_work are waiting on
  5202. getwork before considering the pool lagging.
  5203. - Iterate over each thread belonging to each device in the hashmeter instead of
  5204. searching for them now that they're a list.
  5205. - When using rotate pool strategy, ensure we only select from alive enabled
  5206. pools.
  5207. - Start longpoll from every pool when load balance strategy is in use.
  5208. - Add mandatory and block fields to the work struct. Flag any shares that are
  5209. detected as blocks as mandatory to submit, along with longpoll work from a
  5210. previously rejecting pool.
  5211. - Consider the fan optimal if fanspeed is dropping but within the optimal speed
  5212. window.
  5213. - Fix typo in some API messages (succeess/success)
  5214. - api.c MMQ stat bugs
  5215. - Bugfix: Fix warnings when built without libudev support
  5216. - Bugfix: slay a variety of warnings
  5217. - Bugfix: modminer: Fix unsigned/signed comparison and similar warnings
  5218. - API add ModMinerQuad support
  5219. - Bugfix: Honour forceauto parameter in serial_detect functions
  5220. - modminer: Temperature sensor improvements
  5221. - modminer: Make log messages more consistent in format
  5222. - Only adjust GPU speed up if the fanspeed is within the normal fanrange and
  5223. hasn't been turned to maximum speed under overheat conditions.
  5224. - ModMiner use valid .name
  5225. - New driver: BTCFPGA ModMiner
  5226. - Abstract generally useful FPGA code into fpgautils.c
  5227. - API add stats for pool getworks
  5228. - miner.php option to hide specific fields from the display
  5229. - miner.php add version numbers to the summary page
  5230. - Update debian configs to v2.4.2
  5231. - Add API and FPGA READMEs into Makefile to be included in source distribution.
  5232. - Icarus - fix unit64_t printf warnings
  5233. Version 2.4.2 - June 2, 2012
  5234. - API.class compiled with Java SE 6.0_03 - works with Win7x64
  5235. - miner.php highlight devs too slow finding shares (possibly failing)
  5236. - API update version to V1.11 and document changes
  5237. - API save default config file if none specified
  5238. - api.c save success incorrectly returns error
  5239. - api.c replace BUFSIZ (linux/windows have different values)
  5240. - Move RPC API content out of README to API-README
  5241. - Open a longpoll connection if a pool is in the REJECTING state as it's the
  5242. only way to re-enable it automatically.
  5243. - Use only one longpoll as much as possible by using a pthread conditional
  5244. broadcast that each longpoll thread waits on and checks if it's the current pool
  5245. before
  5246. - If shares are known stale, don't use them to decide to disable a pool for
  5247. sequential rejects.
  5248. - Restarting cgminer from within after ADL has been corrupted only leads to a
  5249. crash. Display a warning only and disable fanspeed monitoring.
  5250. - Icarus: fix abort calculation/allow user specified abort
  5251. - Icarus: make --icarus-timing hidden and document it in FPGA-README
  5252. - Icarus: high accuracy timing and other bitstream speed support
  5253. - add-MIPSEB-to-icarus-for-BIG_ENDIAN
  5254. - work_decode only needs swab32 on midstate under BIG ENDIAN
  5255. - add compile command to api-example.c
  5256. - save config bugfix: writing an extra ',' when no gpus
  5257. - Add dpkg-source commits
  5258. Version 2.4.1 - May 6, 2012
  5259. - In the unlikely event of finding a block, display the block solved count with
  5260. the pool it came from for auditing.
  5261. - Display the device summary on exit even if a device has been disabled.
  5262. - Use correct pool enabled enums in api.c.
  5263. - Import Debian packaging configs
  5264. - Ensure we test for a pool recovering from idle so long as it's not set to
  5265. disabled.
  5266. - Fix pool number display.
  5267. - Give cgminer -T message only if curses is in use.
  5268. - Reinit_adl is no longer used.
  5269. - API 'stats' allow devices to add their own stats also for testing/debug
  5270. - API add getwork stats to cgminer - accesable from API 'stats'
  5271. - Don't initialise variables to zero when in global scope since they're already
  5272. initialised.
  5273. - Get rid of unitialised variable warning when it's false.
  5274. - Move a pool to POOL_REJECTING to be disabled only after 3 minutes of
  5275. continuous rejected shares.
  5276. - Some tweaks to reporting and logging.
  5277. - Change FPGA detection order since BFL hangs on an ICA
  5278. - API support new pool status
  5279. - Add a temporarily disabled state for enabled pools called POOL_REJECTING and
  5280. use the work from each longpoll to help determine when a rejecting pool has
  5281. started working again. Switch pools based on the multipool strategy once a pool
  5282. is re-enabled.
  5283. - Removing extra debug
  5284. - Fix the benchmark feature by bypassing the new networking code.
  5285. - Reset sequential reject counter after a pool is disabled for when it is
  5286. re-enabled.
  5287. - Icarus - correct MH/s and U: with work restart set at 8 seconds
  5288. - ztex updateFreq was always reporting on fpga 0
  5289. - Trying harder to get 1.15y working
  5290. - Specifying threads on multi fpga boards extra cgpu
  5291. - Missing the add cgpu per extra fpga on 1.15y boards
  5292. - API add last share time to each pool
  5293. - Don't try to reap curls if benchmarking is enabled.
  5294. Version 2.4.0 - May 3, 2012
  5295. - Only show longpoll warning once when it has failed.
  5296. - Convert hashes to an unsigned long long as well.
  5297. - Detect pools that have issues represented by endless rejected shares and
  5298. disable them, with a parameter to optionally disable this feature.
  5299. - Bugfix: Use a 64-bit type for hashes_done (miner_thread) since it can overflow
  5300. 32-bit on some FPGAs
  5301. - Implement an older header fix for a label existing before the pthread_cleanup
  5302. macro.
  5303. - Limit the number of curls we recruit on communication failures and with
  5304. delaynet enabled to 5 by maintaining a per-pool curl count, and using a pthread
  5305. conditional that wakes up when one is returned to the ring buffer.
  5306. - Generalise add_pool() functions since they're repeated in add_pool_details.
  5307. - Bugfix: Return failure, rather than quit, if BFwrite fails
  5308. - Disable failing devices such that the user can attempt to re-enable them
  5309. - Bugfix: thread_shutdown shouldn't try to free the device, since it's needed
  5310. afterward
  5311. - API bool's and 1TBS fixes
  5312. - Icarus - minimise code delays and name timer variables
  5313. - api.c V1.9 add 'restart' + redesign 'quit' so thread exits cleanly
  5314. - api.c bug - remove extra ']'s in notify command
  5315. - Increase pool watch interval to 30 seconds.
  5316. - Reap curls that are unused for over a minute. This allows connections to be
  5317. closed, thereby allowing the number of curl handles to always be the minimum
  5318. necessary to not delay networking.
  5319. - Use the ringbuffer of curls from the same pool for submit as well as getwork
  5320. threads. Since the curl handles were already connected to the same pool and are
  5321. immediately available, share submission will not be delayed by getworks.
  5322. - Implement a scaleable networking framework designed to cope with any sized
  5323. network requirements, yet minimise the number of connections being reopened. Do
  5324. this by create a ring buffer linked list of curl handles to be used by getwork,
  5325. recruiting extra handles when none is immediately available.
  5326. - There is no need for the submit and getwork curls to be tied to the pool
  5327. struct.
  5328. - Do not recruit extra connection threads if there have been connection errors
  5329. to the pool in question.
  5330. - We should not retry submitting shares indefinitely or we may end up with a
  5331. huge backlog during network outages, so discard stale shares if we failed to
  5332. submit them and they've become stale in the interim.
  5333. Version 2.3.6 - April 29, 2012
  5334. - Shorten stale share messages slightly.
  5335. - Protect the freeing of current_hash under mutex_lock to prevent racing on it
  5336. when set_curblock is hit concurrently.
  5337. - Change default behaviour to submitting stale, removing the --submit-stale
  5338. option and adding a --no-submit-stale option.
  5339. - Make sure to start the getwork and submit threads when a pool is added on the
  5340. fly. This fixes a crash when a pool is added to running cgminer and then
  5341. switched to.
  5342. - Faster hardware can easily outstrip the speed we can get work and submit
  5343. shares when using only one connection per pool.
  5344. - Test the queued list to see if any get/submits are already queued and if they
  5345. are, start recruiting extra connections by generating new threads.
  5346. - This allows us to reuse network connections at low loads but recuit new open
  5347. connections as they're needed, so that cgminer can scale to hardware of any
  5348. size.
  5349. Version 2.3.5 - April 28, 2012
  5350. - Restarting cgminer leads to a socket that can't be bound for 60 seconds, so
  5351. increase the interval that API binding waits to 30 seconds to minimise the
  5352. number of times it will retry, spamming the logs.
  5353. - Give a longpoll message for any longpoll that detects a block change, primary
  5354. or backup, and also display which pool it was.
  5355. - Decrease utility display to one decimal place.
  5356. - Small cosmetic output alignment.
  5357. - Add pool number to stale share message.
  5358. - Add space to log output now that there is more screen real estate available.
  5359. - Indentation clean up.
  5360. - Merge branch 'master' of github.com:ckolivas/cgminer
  5361. - Remove thread id display from rejected shares as well.
  5362. - Merge pull request #185 from Diapolo/diakgcn
  5363. - add goffset support for diakgcn with -v 1 and update kernel version
  5364. - Set have_longpoll to true when there is at least one pool with longpoll.
  5365. - Don't display the thread ID since it adds no useful information over the
  5366. device number.
  5367. - Don't display the first 8 bytes of a share since they will always be zero at
  5368. >= 1 difficulty.
  5369. - work->longpoll is reset across test_work_current so we need to recheck what
  5370. pool it belongs to.
  5371. - Use longpolls from backup pools with failover-only enabled just to check for
  5372. block changes, but don't use them as work.
  5373. - Start longpoll only after we have tried to extract the longpoll URL.
  5374. - Check for submitold flag on resubmit of shares, and give different message for
  5375. stale shares on retry.
  5376. - Check for submitold before submitstale.
  5377. - Don't force fresh curl connections on anything but longpoll threads.
  5378. - Create one longpoll thread per pool, using backup pools for those pools that
  5379. don't have longpoll.
  5380. - Use the work created from the longpoll return only if we don't have
  5381. failover-enabled, and only flag the work as a longpoll if it is the current
  5382. pool.
  5383. - This will work around the problem of trying to restart the single longpoll
  5384. thread on pool changes that was leading to race conditions.
  5385. - It will also have less work restarts from the multiple longpolls received from
  5386. different pools.
  5387. - Remove the ability to disable longpoll. It is not a useful feature and will
  5388. conflict with planned changes to longpoll code.
  5389. - Remove the invalid entries from the example configuration file.
  5390. - Add support for latest ATI SDK on windows.
  5391. - Export missing function from libztex.
  5392. - miner.php change socktimeoutsec = 10 (it only waits once)
  5393. - Bugfix: Make initial_args a const char** to satisfy exec argument type warning
  5394. (on Windows only)
  5395. - miner.php add a timeout so you don't sit and wait ... forever
  5396. - Create discrete persistent submit and get work threads per pool, thus allowing
  5397. all submitworks belonging to the same pool to reuse the same curl handle, and
  5398. all getworks to reuse their own handle.
  5399. - Use separate handles for submission to not make getwork potentially delay
  5400. share submission which is time critical.
  5401. - This will allow much more reusing of persistent connections instead of opening
  5402. new ones which can flood routers.
  5403. - This mandated a rework of the extra longpoll support (for when pools are
  5404. switched) and this is managed by restarting longpoll cleanly and waiting for a
  5405. thread join.
  5406. - miner.php only show the current date header once
  5407. - miner.php also add current time like single rig page
  5408. - miner.php display rig 'when' table at top of the multi-rig summary page
  5409. - README - add some Ztex details
  5410. - api.c include zTex in the FPGA support list
  5411. - api.c ensure 'devs' shows PGA's when only PGA code is compiled
  5412. - cgminer.c sharelog code consistency and compile warning fix
  5413. - README correct API version number
  5414. - README spelling error
  5415. - api.c combine all pairs of sprintfs()
  5416. - api.c uncomment and use BLANK (and COMMA)
  5417. - Code style cleanup
  5418. - Annotating frequency changes with the changed from value
  5419. - README clarification of 'notify' command
  5420. - README update for API RPC 'devdetails'
  5421. - api.c 'devdetails' list static details of devices
  5422. - Using less heap space as my TP-Link seems to not handle this much
  5423. Version 2.3.4 - April 25, 2012
  5424. - Extensively document the cause of GPU device issues and the use of --gpu-map.
  5425. - Support for share logging
  5426. - Detect poorly performing combination of SDK and phatk kernel and add verbose
  5427. warning at startup.
  5428. - Icarus update to new add_cgpu()
  5429. - Icarus driver working with Linux and Windows
  5430. - api.c fix unused variable compile warning
  5431. - Display all OpenCL devices when -n is called as well to allow debugging of
  5432. differential mapping of OpenCL to ADL.
  5433. - Add a --gpu-map option which will allow arbitrarily mapping ADL devices to
  5434. OpenCL devices for instances where association by enumeration alone fails.
  5435. - Increase upper limit on number of extra items to queue as some FPGA code can't
  5436. yet reliably keep many devices busy.
  5437. - Display configuration file information when -c option is passed and only when
  5438. file exists on loading default config file.
  5439. - Display configuration file loaded, if any, and debug output if configuration
  5440. file parsing failed.
  5441. - Add missing ztex header to Makefile for distribution.
  5442. - Document long-form COM port device names on Windows, required to specify
  5443. serial ports above 9
  5444. - Include ztex bitstreams firmware in distribution and install if configured in.
  5445. - Style police on driver-ztex.c
  5446. - work_restart should only be changed by cgminer.c now
  5447. - Shut down the api cleanly when the api thread is cancelled. This should allow
  5448. the api socket to be closed successfully to next be reopened with app_restart.
  5449. - Make a union for cgpu device handles, and rename "device" to "device_ztex"
  5450. since it's Ztex-specific
  5451. - Initialise name variable.
  5452. - Remove unnecessary check for variable that always has memory allocated.
  5453. - Bugfix: Missing "break" no-op in default case
  5454. - Make the status window and log window as large as can fit on startup,
  5455. rechecking to see if it can be enlarged after the fact. This allows any number
  5456. of devices to be displayed provided the window is made long enough without
  5457. corrupting the output.
  5458. - Style police on libztex.c.
  5459. - API add removepool like the screen interface
  5460. - api.c escape required characters in return strings + pools returns the
  5461. username
  5462. - Set lp_path to NULL after free for consistency.
  5463. - Removing dmalloc import left behind by mistake
  5464. - Fixing leak in resp_hdr_cb
  5465. - miner.php warning highlight GPU stats if they are zero (e.g. ADL not enabled)
  5466. - miner.php highlight any device that isn't 'Enabled'
  5467. - miner.php highlight any Status that isn't 'Alive'
  5468. - miner.php optionally support multiple rigs
  5469. - Initial Ztex support 1.15x board.
  5470. Version 2.3.3 - April 15, 2012
  5471. - Don't even display that cpumining is disabled on ./configure to discourage
  5472. people from enabling it.
  5473. - Do a complete cgminer restart if the ATI Display Library fails, as it does on
  5474. windows after running for some time, when fanspeed reporting fails.
  5475. - Cache the initial arguments passed to cgminer and implement an attempted
  5476. restart option from the settings menu.
  5477. - Disable per-device status lines when there are more than 8 devices since
  5478. screen output will be corrupted, enumerating them to the log output instead at
  5479. startup.
  5480. - Reuse Vals[] array more than W[] till they're re-initialised on the second
  5481. sha256 cycle in poclbm kernel.
  5482. - Minor variable alignment in poclbm kernel.
  5483. - Make sure to disable devices with any status not being DEV_ENABLED to ensure
  5484. that thermal cutoff code works as it was setting the status to DEV_RECOVER.
  5485. - Re-initialising ADL simply made the driver fail since it is corruption over
  5486. time within the windows driver that's responsible. Revert "Attempt to
  5487. re-initialise ADL should a device that previously reported fanspeed stops
  5488. reporting it."
  5489. - Microoptimise poclbm kernel by ordering Val variables according to usage
  5490. frequency.
  5491. Version 2.3.2 - March 31, 2012
  5492. - Damping small changes in hashrate so dramatically has the tendency to always
  5493. make the hashrate underread so go back to gentle damping instead.
  5494. - Revert the crossover of variables from Vals to W in poclbm kernel now that
  5495. Vals are the first declared variables so they're used more frequently.
  5496. - Vals variables appearing first in the array in poclbm is faster.
  5497. - Change the preferred vector width to 1 for Tahiti only, not all poclbm
  5498. kernels.
  5499. - Use a time constant 0.63 for when large changes in hashrate are detected to
  5500. damp change in case the large change is an aliasing artefact instead of a real
  5501. chang
  5502. - Only increment stale counter if the detected stales are discarded.
  5503. - Attempt to re-initialise ADL should a device that previously reported fanspeed
  5504. stops reporting it.
  5505. - Move the ADL setup and clearing to separate functions and provide a reinit_adl
  5506. function to be used when adl fails while running.
  5507. - Use slightly more damping on the decay time function in the never-ending quest
  5508. to smooth off the hashmeter.
  5509. - Set the starting fanspeed to a safe and fairly neutral 50% when autofan is
  5510. enabled.
  5511. - Provide locking around updates of cgpu hashrates as well to prevent multiple
  5512. threads accessing data fields on the same device.
  5513. - Display the beginning of the new block in verbose mode in the logs.
  5514. - Reinstate old diablo kernel variable ordering from 120222, adding only goffset
  5515. and vector size hint. The massive variable ordering change only helped one SDK
  5516. on
  5517. - Change the version number on the correct kernels.
  5518. - api.c devicecode/osinfo incorrectly swapped for json
  5519. - Add extensive instructions on how to make a native windows build.
  5520. - Update version numbers of poclbm and diablo kernels as their APIs have also
  5521. changed.
  5522. - Use global offset parameter to diablo and poclbm kernel ONLY for 1 vector
  5523. kernels.
  5524. - Use poclbm preferentially on Tahiti now regardless of SDK.
  5525. - Remove unused constant passed to poclbm.
  5526. - Clean up use of macros in poclbm and use bitselect everywhere possible.
  5527. - Add vector type hint to diablo kernel.
  5528. - Add worksize and vector attribute hints to the poclbm kernel.
  5529. - Spaces for non-aligned variables in poclbm.
  5530. - More tidying of poclbm.
  5531. - Swap Vals and W variables where they can overlap in poclbm.
  5532. - More tidying of poclbm.
  5533. - Tidy up first half of poclbm.
  5534. - Clean up use of any() by diablo and poclbm kernels.
  5535. - Minor variable symmetry changes in poclbm.
  5536. - Put additions on separate lines for consistency in poclbm.
  5537. - Consolidate last use of W11 into Vals4 in poclbm.
  5538. - Change email due to SPAM
  5539. - api.c miner.php add a '*' to the front of all notify counters - simplifies
  5540. future support of new counters
  5541. - miner.php add display 'notify' command
  5542. - Small change to help arch's without processor affinity
  5543. - Fix bitforce compile error
  5544. - api.c notify should report disabled devices also - of course
  5545. - API returns the simple device history with the 'notify' command
  5546. - code changes for supporting a simple device history
  5547. - api.c Report an OS string in config to help with device issues
  5548. - api.c fix Log Interval - integer in JSON
  5549. - api.c config 'Device Code' to show list of compiled devices + README
  5550. - api.c increase buffer size close to current code allowable limit
  5551. - removed 8-component vector support from kernel, as this is not supported in
  5552. CGMINER anyway
  5553. - forgot to update kernel modification date, fixed ;)
  5554. - reordered an addition in the kernel, which results in less instructions used
  5555. in the GPU ISA code for GCN
  5556. - miner.php: option for readonly or check privileged access
  5557. - Ignore reduntant-with-build options --disable-gpu, --no-adl, and --no-restart
  5558. - miner.php: ereg_replace is DEPRECATED so use preg_replace instead
  5559. - Make curses TUI support optional at compile-time.
  5560. - Bugfix: AC_ARG_WITH provides withval instead of enableval
  5561. - miner.php split devs output for different devices
  5562. - api.c: correct error messages
  5563. - icarus.c modify (regular) timeout warning to only be debug
  5564. - icarus.c set the windows TODO timeout
  5565. - Allow specifying a specific driver for --scan-serial
  5566. - optimized nonce-check and output code for -v 2 and -v 4
  5567. - Bugfix: Check for libudev header (not just library) in configure, and document
  5568. optional dependency
  5569. - Add API support for Icarus and Bitforce
  5570. - Next API version is 1.4 (1.3 is current)
  5571. - README/api.c add "When" the request was processed to STATUS
  5572. - Bugfix: ZLX to read BitFORCE temp, not ZKX -.-
  5573. - Use libudev to autodetect BitFORCE GPUs, if available
  5574. - Use the return value of fan_autotune to set fan_optimal instead of passing it
  5575. as a pointer.
  5576. - Pass the lasttemp from the device we're using to adjust fanspeed in twin
  5577. devices.
  5578. - fix the name to 3 chars, fix the multi-icarus support
  5579. - Bugfix: "-S auto" is the default if no -S is specified, and there is no such
  5580. delay in using it
  5581. - README add information missing from --scan-serial
  5582. - Update README RPC API Version comment
  5583. - Bugfix: Allow enabling CPU even without OpenCL support
  5584. - Change failed-to-mine number of requested shares messge to avoid segfault on
  5585. recursive calling of quit().
  5586. - Get rid of extra char which is just truncated in poclbm kernel.
  5587. - only small code formating changes
  5588. - removed vec_step() as this could lead to errors on older SDKs
  5589. - unified code for generating nonce in kernel and moved addition of base to the
  5590. end -> faster
  5591. Version 2.3.1 - February 24, 2012
  5592. - Revert input and output code on diakgcn and phatk kernels to old style which
  5593. worked better for older hardware and SDKs.
  5594. - Add a vector*worksize parameter passed to those kernels to avoid one op.
  5595. - Increase the speed of hashrate adaptation.
  5596. - Only send out extra longpoll requests if we want longpolls.
  5597. - API implement addpool command
  5598. - API return the untouched Total MH also (API now version 1.3)
  5599. - Add enable/disablepool to miner.php example and reduce font size 1pt
  5600. Version 2.3.0 - February 23, 2012
  5601. - Consider extra longpoll work items as staged_extra so as to make sure we queue
  5602. more work if queueing regular work items as longpolls.
  5603. - Use diablo kernel on all future SDKs for Tahiti and set preferred vector width
  5604. to 1 on poclbm kernel only.
  5605. - Explicitly type the constants in diakgcn kernel as uint, to be in line with
  5606. poclbm kernel.
  5607. - Reset all hash counters at the same time as resetting start times to get
  5608. accurate hashrates on exiting which is mandatory for benchmarking.
  5609. - Report thread out before it starts to avoid being flagged as sick when waiting
  5610. for the first work item.
  5611. - Don't disable and re-enable devices as they may recover and in the meantime
  5612. have their status set to OFF.
  5613. - API new commands enablepool and disablepool (version already incremented)
  5614. - Tolerate new-format temperature readings for bitforce
  5615. - Modify cgminer.c pool control to allow API to call it
  5616. - Bugfix: Fix BitFORCE driver memory leak in debug logging
  5617. - Extra byte was being unused in poclbm leading to failure on some platforms.
  5618. - Explicitly type the constants in poclbm kernel as uint.
  5619. - Don't save 'include' when saving the configuration
  5620. - Allow configuration file to include another recursively
  5621. - Use the SDK and hardware information to choose good performing default
  5622. kernels.
  5623. - Move phatk kernel to offset vector based nonce bases as well.
  5624. - Add a --benchmark feature which works on a fake item indefinitely to compare
  5625. device performance without any server or networking influence.
  5626. - Allow writing of multiple worksizes to the configuration file.
  5627. - Allow writing of multiple vector sizes to the configuration file.
  5628. - Allow writing of multiple kernels to the configuration file.
  5629. - Allow multiple different kernels to be chosen per device.
  5630. - Allow the worksize to be set per-device.
  5631. - Allow different vectors to be set per device.
  5632. - If we're well below the target temperature, increase gpu engine speed back to
  5633. maximum in case we have gotten lost between profiles during an idle period.
  5634. - We should be setting the value of fan_optimal, not its address.
  5635. - As all kernels will be new versions it's an opportunity to change the .bin
  5636. format and make it simpler. Specifying bitalign is redundant and long can be l.
  5637. - Use any() in kernel output code.
  5638. - Put the nonce for each vector offset in advance, avoiding one extra addition
  5639. in the kernel.
  5640. - Reset times after all mining threads are started to make estimating hashrates
  5641. easier at startup.
  5642. - Bugfix: allow no-exec (NX) stack
  5643. - Fix minor warning.
  5644. - fix the bitforce.c code style follow 1TBS
  5645. - fix icarus.c compile warning
  5646. - small changes to speedup no vec for AMD 898.1 OCL runtime
  5647. - Update licensing to GPL V3.
  5648. - Reset the longpoll flag after it's been used once to prevent it restarting
  5649. work again.
  5650. - Begin import of DiabloMiner kernel.
  5651. - Modify API debug messages to say API instead of DBG
  5652. - When API shuts down cgminer don't kill itself
  5653. - Don't make rolled work from the longpoll be seen as other longpoll work items.
  5654. - API add 'privileged' command so can verify access level
  5655. - Set the lp_sent variable under lock since there will almost always be a race
  5656. on setting this variable, potentially leading to multiple LPs being sent out.
  5657. - API restrict access to all non display commands by default
  5658. - Update API version to 1.2 for new 'Log Interval'
  5659. - API add --log Interval to 'config' reply
  5660. - --api-allow special case 0/0 means all
  5661. Version 2.2.7 - February 20, 2012
  5662. - Send out extra longpolls when we have switched pools and the longpoll thread
  5663. is still bound to the old one. This is particularly useful with p2pool where
  5664. longpolls do not correlate with main bitcoin block change and would have led to
  5665. high reject rates on failover.
  5666. - Store whether a work item is the result of a longpoll or not in struct work
  5667. and use it to help determine block changes directly from the work longpoll bool.
  5668. - Keep track of when a longpoll has been sent for a pool and if the current pool
  5669. is requesting work but has not sent a longpoll request, convert one of the work
  5670. items to a longpoll.
  5671. - Store the longpoll url in the pool struct and update it from the pool_active
  5672. test in case it changes. This is to allow further changes to longpoll management
  5673. on switching pools.
  5674. - Re-check for a longpoll supporting pool every 30 seconds if none is found
  5675. initially.
  5676. - Report threads as busy waiting on getwork on startup to avoid them being
  5677. flagged sick on startup during slow networking.
  5678. - Allow devices that are disabled due to overheating to be flagged as recovering
  5679. instead of disabling them and re-enable them if they're below ideal temperatures
  5680. - Tahiti prefers worksize 64 with poclbm.
  5681. - No need to expressly retain the opencl program now that the zero binary issue
  5682. is fixed. This actually fixes cgminer to work with the latest SDK included with
  5683. the ATI catalyst driver 12.2.
  5684. - Show error code on any opencl failure status.
  5685. - Add detection for version 898.1 SDK as well but only give SDK 2.6 warning once
  5686. on startup instead of with each device initialisation.
  5687. - Always use a fresh connection for longpoll as prolonged persistent connections
  5688. can fail for many reasons.
  5689. - Keep track of intended engine clock speed and only adjust up if it's higher
  5690. than the last intended speed. This avoids setting the clock speed to one
  5691. relative to a lower profile one by mistake.
  5692. - Use gpu-memdiff on startup if an engine clockspeed is set and a memdiff value
  5693. is set.
  5694. - Revert "Adjust engine speed up according to performance level engine setting,
  5695. not the current engine speed." - ineffectual.
  5696. - Freeze the queues on all threads that are sent the pause message to prevent
  5697. them trying to start up again with saved pings in their queues.
  5698. - Updates to diakgcn kernel/
  5699. - Consolidate all screen updates to the watchdog thread and touch both windows
  5700. before refresh.
  5701. - Curses will be disabled in clean_up so don't do it early in kill_work, and
  5702. disable_adl so that GPU settings may be restored to normal in case shutting down
  5703. curses leads to instability on windows.
  5704. - Stop the mining threads before trying to kill them.
  5705. - Plain refresh() does not give reliably screen updates so get rid of all uses
  5706. of it.
  5707. - First release with working diakgcn kernel.
  5708. Version 2.2.6 - February 16, 2012
  5709. - Provide warning on each startup about sdk 2.6
  5710. - Fix unused warnings on win32.
  5711. - bitforce: Simplify BFopen WIN32 ifdef/else
  5712. - Fix initialization warning with jansson 1.3
  5713. - bitforce: Cleanup extraneous TODO that isn't needed
  5714. - Move tcsetattr (and new tcflush) into *nix BFopen to simplify things a bit
  5715. - Add message explaining 2nd thread disabling for dynamic mode and how to tune
  5716. it.
  5717. - Move logwindow down once number of devices is known.
  5718. - Automatically choose phatk kernel for bitalign non-gcn ATI cards, and then
  5719. only select poclbm if SDK2.6 is detected.
  5720. - Allow the refresh interval to be adjusted in dynamic intensity with a
  5721. --gpu-dyninterval parameter.
  5722. - Make curses display visible right from the beginning and fix the window sizes
  5723. so the initial messages don't get lost once the status window is drawn.
  5724. - The amount of work scanned can fluctuate when intensity changes and since we
  5725. do this one cycle behind, we increment the work more than enough to prevent
  5726. repeati
  5727. - bitforce: Set a 30 second timeout for serial port on Windows, since the
  5728. default is undefined
  5729. - Use PreVal4addT1 instead of PreVal4 in poclbm kernel.
  5730. - Import PreVal4 and PreVal0 into poclbm kernel.
  5731. - Import more prepared constants into poclbm kernel.
  5732. - Keep variables in one array but use Vals[] name for consistency with other
  5733. kernel designs.
  5734. - Replace constants that are mandatorily added in poclbm kernel with one value.
  5735. - Remove addition of final constant before testing for result in poclbm kernel.
  5736. - Hand optimise variable addition order.
  5737. - Hand optimise first variable declaration order in poclbm kernel.
  5738. - Radical reordering machine based first pass to change variables as late as
  5739. possible, bringing their usage close together.
  5740. - fix strcpy NULL pointer if env HOME unset.
  5741. - bitforce: Disable automatic scanning when at least one device is specified
  5742. manually
  5743. - Unroll all poclbm additions to enable further optimisations.
  5744. Version 2.2.5 - February 13, 2012
  5745. - Make output buffer write only as per Diapolo's suggestion.
  5746. - Constify nonce in poclbm.
  5747. - Use local and group id on poclbm kernel as well.
  5748. - Microoptimise phatk kernel on return code.
  5749. - Adjust engine speed up according to performance level engine setting, not the
  5750. current engine speed.
  5751. - Try to load a binary if we've defaulted to the poclbm kernel on SDK2.6
  5752. - Use the poclbm kernel on SDK2.6 with bitalign devices only if there is no
  5753. binary available.
  5754. - Further generic microoptimisations to poclbm kernel.
  5755. - The longstanding generation of a zero sized binary appears to be due to the
  5756. OpenCL library putting the binary in a RANDOM SLOT amongst 4 possible binary
  5757. locations. Iterate over each of them after building from source till the real
  5758. binary is found and use that.
  5759. - Fix harmless warnings with -Wsign-compare to allow cgminer to build with -W.
  5760. - Fix missing field initialisers warnings.
  5761. - Put win32 equivalents of nanosleep and sleep into compat.h fixing sleep() for
  5762. adl.c.
  5763. - Restore compatibility with Jansson 1.3 and 2.0 (api.c required 2.1)
  5764. - Modularized logging, support for priority based logging
  5765. - Move CPU chipset specific optimization into device-cpu
  5766. Version 2.2.4 - February 11, 2012
  5767. - Fix double definition of A0 B0 to zeroA zeroB.
  5768. - Retain cl program after successfully loading a binary image. May decrease
  5769. failures to build kernels at startup.
  5770. - Variable unused after this so remove setting it.
  5771. - BFI INT patching is not necessarily true on binary loading of files and not
  5772. true on ATI SDK2.6+. Report bitalign instead.
  5773. - Various string fixes for reject reason.
  5774. - Generalize --temp-cutoff and implement support for reading temperature from
  5775. BitFORCE FPGAs
  5776. - Change message from recovered to alive since it is used on startup as well as
  5777. when a pool has recovered.
  5778. - Start mining as soon as any pool is found active and rely on the watchpool
  5779. thread to bring up other pools.
  5780. - Delayed responses from testing pools that are down can hold up the watchdog
  5781. thread from getting to its device testing code, leading to false detection of
  5782. the GPU not checking in, and can substantially delay auto gpu/auto fan
  5783. management leading to overheating. Move pool watching to its own thread.
  5784. - Bugfix: BitFORCE index needs to be static to count correctly
  5785. - Space out retrieval of extra work according to the number of mining threads.
  5786. - Make shutdown more robust. Enable the input thread only after the other
  5787. threads exist. Don't kill off the workio thread and use it to exit main() only
  5788. if there is an unexpected problem. Use kill_work() for all anticipated shutdowns
  5789. where possible. Remove unused thread entry.
  5790. - Change poclbm version number.
  5791. - One array is faster than 2 separate arrays so change to that in poclbm kernel.
  5792. - Microoptimisations to poclbm kernel which increase throughput slightly.
  5793. - Import diablominer kernel. Currently disabled as not working.
  5794. - Import diapolo kernel. Currently disabled as not working.
  5795. - Conflicting entries of cl_kernel may have been causing problems, and
  5796. automatically chosen kernel type was not being passed on. Rename the enum to
  5797. cl_kernels and store the chosen kernel in each clState.
  5798. - Set cl_amd_media_ops with the BITALIGN flag and allow non-bitselect devices to
  5799. build.
  5800. - ALlow much longer filenames for kernels to load properly.
  5801. - Allow different kernels to be used by different devices and fix the logic fail
  5802. of overcorrecting on last commit with !strstr.
  5803. - Fix kernel selection process and build error.
  5804. - queue_phatk_kernel now uses CL_SET_VARG() for base-nonce(s), too
  5805. - added OpenCL >= 1.1 detection code, in preparation of OpenCL 1.1 global offset
  5806. parameter support
  5807. - Use K array explicitly to make it clear what is being added.
  5808. - Work items have a tendency to expire at exactly the same time and we don't
  5809. queue extra items when there are plenty in the queue, regardless of age. Allow
  5810. extra work items to be queued if adequate time has passed since we last
  5811. requested work even if over the limit.
  5812. - Discard work when failover-only is enabled and the work has come from a
  5813. different pool.
  5814. - Missing include to build on newer mingw32.
  5815. - Move from the thread safe localtime_r to regular localtime which is the only
  5816. one supported on newer pthread libraries on mingw32 to make it compile with the
  5817. newer ming. Thread safety is of no importance where localtime is used in this
  5818. code.
  5819. - Define in_addr_t in windows if required
  5820. - sys/wait.h not required in windows
  5821. - Allow API to restrict access by IP address
  5822. - Add pool switching to example miner.php
  5823. - Display X-Reject-Reason, when provided
  5824. - Remove the test for whether the device is on the highest profil level before
  5825. raising the GPU speed as it is ineffectual and may prevent raising the GPU
  5826. speed.
  5827. - Remove unnecessary check for opt_debug one every invocation of applog at
  5828. LOG_DEBUG level and place the check in applog().
  5829. Version 2.2.3 - February 6, 2012
  5830. - Revert "Rewrite the convoluted get_work() function to be much simpler and roll
  5831. work as much as possible with each new work item." This seems to cause a race on
  5832. work in free_work(). Presumably other threads are still accessing the structure.
  5833. Version 2.2.2 - February 6, 2012
  5834. - Provide support for the submitold extension on a per-pool basis based on the
  5835. value being detected in a longpoll.
  5836. - Don't send a ping to a dynamic device if it's not enabled as that will just
  5837. enable it for one pass and then disable it again.
  5838. - Rewrite the convoluted get_work() function to be much simpler and roll work as
  5839. much as possible with each new work item.
  5840. - Roll as much work as possible from the work returned from a longpoll.
  5841. - Rolling work on each loop through the mining thread serves no purpose.
  5842. - Allow to stage more than necessary work items if we're just rolling work.
  5843. - Replace divide_work with reuse_work function used twice.
  5844. - Give rolled work a new ID to make sure there is no confusion in the hashtable
  5845. lookups.
  5846. - Remove now-defunct hash_div variables.
  5847. - Remove unused get_dondata function.
  5848. - Silence ADL warnings.
  5849. - Silence unused parameter warnings.
  5850. - Stagger the restart of every next thread per device to keep devices busy ahead
  5851. of accessory threads per device.
  5852. - Deprecate the --donation feature. Needlessly complex, questionable usefulness,
  5853. depends on author's server and a central pool of some kind, and was not heavily
  5854. adopted.
  5855. - It's devices that report back now, not threads, update message.
  5856. - Continue auto-management of fan and engine speeds even if a device is disabled
  5857. for safety reasons.
  5858. - No need to check we're highest performance level when throttling GPU engine
  5859. speed.
  5860. - Abstract out tests for whether work has come from a block that has been seen
  5861. before and whether a string is from a previously seen block.
  5862. - Probe but don't set the timeout to 15 seconds as some networks take a long
  5863. time to timeout.
  5864. - Remove most compiler warnings from api.c
  5865. - Add last share's pool info in cgpu_info
  5866. - Allow the OpenCL platform ID to be chosen with --gpu-platform.
  5867. - Iterate over all platforms displaying their information and number of devices
  5868. when --ndevs is called.
  5869. - Deprecate main.c
  5870. - Some networks can take a long time to resolve so go back to 60 second timeouts
  5871. instead of 15.
  5872. - Only enable curses on failure if curses is desired.
  5873. - Fix warnings in bitforce.c
  5874. - Bugfix: Need to open BitForce tty for read-write
  5875. - Fix various build issues.
  5876. - Modularize code: main.c -> device-cpu + device-gpu
  5877. - Fix phatk kernel not working on non-bitalign capable devices (Nvidia, older
  5878. ATI).
  5879. - Update poclbm kernel for better performance on GCN and new SDKs with bitalign
  5880. support when not BFI INT patching. Update phatk kernel to work properly for non
  5881. BFI INT patched kernels, providing support for phatk to run on GCN and non-ATI
  5882. cards.
  5883. - Return last accepted share pool/time for devices
  5884. - Display accepted share pool/time for CPUs
  5885. - Bug intensity always shows GPU 0
  5886. - Update example web miner.php to use new API commands
  5887. Version 2.2.1 - January 30, 2012
  5888. NOTE - The GPU Device reordering in 2.2.0 by default was considered a bad idea
  5889. so the original GPU ordering is used by default again unless reordering is
  5890. explicitly requested.
  5891. - Fix bitforce failing to build into cgminer.
  5892. - Add missing options to write config function.
  5893. - Add a --gpu-reorder option to only reorder devices according to PCI Bus ID
  5894. when requested.
  5895. - Fix for midstate support being broken on pools that supported no-midstate
  5896. work by ensuring numbers are 32 bits in sha2.c
  5897. - Set virtual GPUs to work when ADL is disabled or all mining will occur on GPU
  5898. 0.
  5899. - Add information about paused threads in the menu status.
  5900. - Disable all but the first thread on GPUs in dynamic mode for better
  5901. interactivity.
  5902. - Set the latest network access time on share submission for --net-delay even if
  5903. we're not delaying that submission for further network access.
  5904. - Clear adl on exiting after probing values since it may attempt to overclock.
  5905. - As share submission is usually staggered, and delays can be costly, submit
  5906. shares without delay even when --net-delay is enabled.
  5907. - Display GPU number and device name when ADL is successfully enabled on it.
  5908. - Display GPU ordering remapping in verbose mode.
  5909. - Don't fail in the case the number of ADL and OpenCL devices do not match, and
  5910. do not attempt to reorder devices unless they match. Instead give a warning
  5911. about
  5912. - Display error codes should ADL not return ADL_OK in the more critical function
  5913. calls.
  5914. - Fix unused warning.
  5915. - Fix compile warnings in api.c
  5916. - Add extensive ADL based device info in debug mode.
  5917. - Make --ndevs display verbose opencl information as well to make debugging
  5918. version information easier.
  5919. - Display information about the opencl platform with verbose enabled.
  5920. - Explicitly check for nvidia in opencl platform strings as well.
  5921. Version 2.2.0 - January 29, 2012
  5922. NOTE: GPU Device order will change with this release with ATI GPUs as cgminer
  5923. now can enumerate them according to their Bus ID which means the values should
  5924. now correlate with their physical position on the motherboard.
  5925. - Default to poclbm kernel on Tahiti (7970) since phatk does not work, even
  5926. though performance is sub-standard so that at least it will mine successfully by
  5927. defau
  5928. - Retain cl program after every possible place we might build the program.
  5929. - Update ADL SDK URL.
  5930. - Fix potential overflow.
  5931. - Map GPU devices to virtual devices in their true physical order based on
  5932. BusNumber.
  5933. - Change the warning that comes with failure to init cl on a device to be more
  5934. generic and accurate.
  5935. - Advertise longpoll support in X-Mining-Extensions
  5936. - Detect dual GPU cards by iterating through all GPUs, finding ones without
  5937. fanspeed and matching twins with fanspeed one bus ID apart.
  5938. - Do not attempt to build the program that becomes the kernel twice. This could
  5939. have been leading to failures on initialising cl.
  5940. - Some opencl compilers have issues with no spaces after -D in the compiler
  5941. options.
  5942. - Allow intensity up to 14.
  5943. - Use calloced stack memory for CompilerOptions to ensure sprintf writes to the
  5944. beginning of the char.
  5945. - Whitelist 79x0 cards to prefer no vectors as they perform better without.
  5946. - Adjust fan speed gently while in the optimal range when temperature is
  5947. drifting to minimise overshoot in either direction.
  5948. - Detect dual GPU cards via the indirect information of - 1st card has a fan
  5949. controller. 2nd card does not have a fan controller, cards share the same device
  5950. name
  5951. - Instead of using the BFI_INT patching hack on any device reporting
  5952. cl_amd_media_ops, create a whitelist of devices that need it. This should enable
  5953. GCN architec
  5954. - Fixed API compiling issue on OS X
  5955. - Add more explanation of JSON format and the 'save' command
  5956. - Return an error if using ADL API commands when it's not available
  5957. - Read off lpThermalControllerInfo from each ADL device.
  5958. - Add ADL_Overdrive5_ThermalDevices_Enum interface.
  5959. - Add API commands: config, switchpool, gpu settings, save
  5960. - Implement socks4 proxy support.
  5961. - Fix send() for JSON strings
  5962. - Introduce a --net-delay option which guarantees at least 250ms between any
  5963. networking requests to not overload slow routers.
  5964. - Generalise locking init code.
  5965. - Allow invalid values to be in the configuration file, just skipping over them
  5966. provided the rest of the file is valid JSON. This will allow older configurat
  5967. - Allow CPU mining explicitly enable only if other mining support is built in.
  5968. - BitForce FPGA support
  5969. - Configure out building and support of all CPU mining code unless
  5970. --enable-cpumining is enabled.
  5971. - Allow parsed values to be zero which will allow 0 values in the config file to
  5972. work.
  5973. - Advertise that we can make our own midstate, so the pool can skip generating
  5974. it for us
  5975. - Refactor the CPU scanhash_* functions to use a common API. Fixes bugs.
  5976. - Don't consider a pool lagging if a request has only just been filed. This
  5977. should decrease the false positives for "pool not providing work fast enough".
  5978. - Invalidating work after longpoll made hash_pop return no work giving a false
  5979. positive for dead pool. Rework hash_pop to retry while finds no staged work u
  5980. - Remove TCP_NODELAY from curl options as many small packets may be contributing
  5981. to network overload, when --net-delay is enabled.
  5982. - Refactor miner_thread to be common code for any kind of device
  5983. - Simplify submit_nonce loop and avoid potentially missing FOUND - 1 entry.
  5984. Reported by Luke-Jr.
  5985. - Micro-optimisation in sha256_sse2 code courtesy of Guido Ascioti
  5986. guido.ascioti@gmail.com
  5987. - Refactor to abstract device-specific code
  5988. Version 2.1.2 - January 6, 2012
  5989. - If api-description is specified, save it when writing the config file
  5990. - Adjust utility width to be constant maximum as well.
  5991. - Add percent signs to reject ratio outputs
  5992. - Should the donation pool fail, don't make the fallover pool behave as though
  5993. the primary pool is lagging.
  5994. - Use an alternative pool should the donation getwork fail.
  5995. Version 2.1.1 - January 1, 2012
  5996. - Include API examples in distribution tarball.
  5997. - Don't attempt to pthread_join when cancelling threads as they're already
  5998. detached and doing so can lead to a segfault.
  5999. - Give more generic message if slow pool at startup is the donation pool.
  6000. - Continue to attempt restarting GPU threads if they're flagged dead at 1 min.
  6001. intervals.
  6002. - Don't attempt to restart sick flagged GPUs while they're still registering
  6003. activity.
  6004. - Make curl use fresh connections whenever there is any communication issue
  6005. in case there are dead persistent connections preventing further comms from
  6006. working.
  6007. - Display pool in summary if only 1 pool.
  6008. - Adjust column width of A/R/HW to be the maximum of any device and align them.
  6009. Version 2.1.0 - December 27, 2011
  6010. - Major infrastructure upgrade with RPC interface for controlling via sockets
  6011. encoded with/without JSON courtesy of Andrew Smith. Added documentation for
  6012. use of the API and sample code to use with it.
  6013. - Updated linux-usb-cgminer document.
  6014. - Rewrite of longpoll mechanism to choose the current pool wherever possible to
  6015. use for the longpoll, or any pool that supports longpoll if the current one
  6016. does not.
  6017. - Display information about longpoll when the chosen server has changed.
  6018. - Fix the bug where longpoll generated work may have been sent back to the
  6019. wrong pool, causing rejects.
  6020. - Fix a few race conditions on closing cgminer which caused some of the crashes
  6021. on exit.
  6022. - Only adjust gpu engine speed in autotune mode if the gpu is currently at the
  6023. performance level of that being adjusted.
  6024. - Various fixes for parsing/writing of configuration files.
  6025. - Do not add blank lines for threads of unused CPUs.
  6026. - Show which pool is unresponsive on startup.
  6027. - Only show GPU management menu item if GPUs are in use.
  6028. - Align most device columns in the curses display.
  6029. Version 2.0.8 - November 11, 2011
  6030. - Make longpoll do a mandatory flushing of all work even if the block hasn't
  6031. changed, thus supporting longpoll initiated work change of any sort and merged
  6032. mining.
  6033. - Byteswap computed hash in hashtest so it can be correctly checked. This fixes
  6034. the very rare possibility that a block solve on solo mining was missed.
  6035. - Add x86_64 w64 mingw32 target
  6036. - Allow a fixed speed difference between memory and GPU clock speed with
  6037. --gpu-memdiff that will change memory speed when GPU speed is changed in
  6038. autotune mode.
  6039. - Don't load the default config if a config file is specified on the command
  6040. line.
  6041. - Don't build VIA on apple since -a auto bombs instead of gracefully ignoring
  6042. VIA failing.
  6043. - Build fix for dlopen/dlclose errors in glibc.
  6044. Version 2.0.7 - October 17, 2011
  6045. - Support work without midstate or hash1, which are deprecated in bitcoind 0.5+
  6046. - Go to kernel build should we fail to clCreateProgramWithBinary instead of
  6047. failing on that device. This should fix the windows problems with devices not
  6048. initialising.
  6049. - Support new configuration file format courtesy of Chris Savery which can write
  6050. the config file from the menu and will load it on startup.
  6051. - Write unix configuration to .cgminer/cgminer.conf by default and prompt to
  6052. overwrite if given a filename from the menu that exists.
  6053. Version 2.0.6 - October 9, 2011
  6054. - Must initialise the donorpool mutex or it fails on windows.
  6055. - Don't make donation work interfere with block change detection allowing
  6056. donation to work regardless of the block chain we're mining on.
  6057. - Expire shares as stale with a separate timeout from the scantime, defaulting
  6058. to 120 seconds.
  6059. - Retry pools after a delay of 15 seconds if none can be contacted on startup
  6060. unless a key is pressed.
  6061. - Don't try to build adl features without having adl.
  6062. - Properly check shares against target difficulty - This will no longer show
  6063. shares when solo mining at all unless they're considered to be a block solve.
  6064. - Add altivec 4 way (cpu mining) support courtesy of Gilles Risch.
  6065. - Try to use SSL if the server supports it.
  6066. - Display the total solved blocks on exit (LOL if you're lucky).
  6067. - Use ADL activity report to tell us if a sick GPU is still busy suggesting it
  6068. is hard hung and do not attempt to restart it.
  6069. Version 2.0.5 - September 27, 2011
  6070. - Intensity can now be set to dynamic or static values per-device.
  6071. - New donation feature --donation sends a proportion of shares to author's
  6072. account of choice, but is disabled by default!
  6073. - The hash being displayed and block detection has been fixed.
  6074. - Devices not being mined on will not attempt to be ADL managed.
  6075. - Intensity is now displayed per GPU device.
  6076. - Make longpoll attempt to restart as often as opt_retries specifies.
  6077. - We weren't rolling work as often as we could.
  6078. - Correct some memory management issues.
  6079. - Build fixes.
  6080. - Don't mess with GPUs if we don't have them.
  6081. Version 2.0.4 - September 23, 2011
  6082. - Confused Longpoll messages should be finally fixed with cgminer knowing for
  6083. sure who found the new block and possibly avoiding a rare crash.
  6084. - Display now shows the actual hash and will say BLOCK! if a block is deemed
  6085. solved.
  6086. - Extra spaces, which would double space lines on small terminals, have been
  6087. removed.
  6088. - Fan speed change is now damped if it is already heading in the correct
  6089. direction to minimise overshoot.
  6090. - Building without opencl libraries is fixed.
  6091. - GPUs are autoselected if there is only one when in the GPU management menu.
  6092. - GPU menu is refreshed instead of returning to status after a GPU change.
  6093. Version 2.0.3 - September 17, 2011
  6094. - Various modes of failure to set fanspeeds and adl values have been addressed
  6095. and auto-fan should work now on most hardware, and possibly other values
  6096. which previously would not have worked.
  6097. - Fixed a crash that can occur on switching pools due to longpoll thread races.
  6098. - Use ATISTREAMSDKROOT if available at build time.
  6099. - Fanspeed management is returned to the driver default on exit instead of
  6100. whatever it was when cgminer was started.
  6101. - Logging of events deemed WARNING or ERR now will display even during
  6102. periods where menu input is being awaited on.
  6103. Version 2.0.2 - September 11, 2011
  6104. - Exit cleanly if we abort before various threads are set up or if they no
  6105. longer exist.
  6106. - Fix a rare crash in HASH_DEL due to using different mutexes to protect the
  6107. data.
  6108. - Flag devices that have never started and don't allow enabling of devices
  6109. without restarting them.
  6110. - Only force the adapter speed to high if we've flagged this device as being
  6111. managed.
  6112. - Flag any devices with autofan or autogpu as being managed.
  6113. - Use a re-entrant value to store what fanspeed we're trying to set in case the
  6114. card doesn't support small changes. Force it to a multiple of 10% if it
  6115. fails on trying to speed up the fan.
  6116. - Do not bother resetting values to old ones if changes to GPU parameters report
  6117. failure, instead returning a failure code only if the return value from get()
  6118. differs.
  6119. - Remove redundant check.
  6120. - Only display supported values from fanspeed on change settings.
  6121. - Missing bracket from output.
  6122. - Display fan percentage on devices that only support reporting percent and not
  6123. RPM.
  6124. - Properly substitute DLOPEN flags to build with ADL support when -ldl is needed
  6125. and not when opencl is not found.
  6126. Version 2.0.1 - September 9, 2011
  6127. - Fix building on 32bit glibc with dlopen with -lpthread and -ldl
  6128. - ByteReverse is not used and the bswap opcode breaks big endian builds. Remove
  6129. it.
  6130. - Ignore whether the display is active or not since only display enabled devices
  6131. work this way, and we skip over repeat entries anwyay.
  6132. - Only reset values on exiting if we've ever modified them.
  6133. - Flag adl as active if any card is successfully activated.
  6134. - Add a thermal cutoff option as well and set it to 95 degrees by default.
  6135. - Change the fan speed by only 5% if it's over the target temperature but less
  6136. than the hysteresis value to minimise overshoot down in temperature.
  6137. - Add a --no-adl option to disable ADL monitoring and GPU settings.
  6138. - Only show longpoll received delayed message at verbose level.
  6139. - Allow temperatures greater than 100 degrees.
  6140. - We should be passing a float for the remainder of the vddc values.
  6141. - Implement accepting a range of engine speeds as well to allow a lower limit to
  6142. be specified on the command line.
  6143. - Allow per-device fan ranges to be set and use them in auto-fan mode.
  6144. - Display which GPU has overheated in warning message.
  6145. - Allow temperature targets to be set on a per-card basis on the command line.
  6146. - Display fan range in autofan status.
  6147. - Setting the hysteresis is unlikely to be useful on the fly and doesn't belong
  6148. in the per-gpu submenu.
  6149. - With many cards, the GPU summaries can be quite long so use a terse output
  6150. line when showing them all.
  6151. - Use a terser device status line to show fan RPM as well when available.
  6152. - Define max gpudevices in one macro.
  6153. - Allow adapterid 0 cards to enumerate as a device as they will be non-AMD
  6154. cards, and enable ADL on any AMD card.
  6155. - Do away with the increasingly confusing and irrelevant total queued and
  6156. efficiency measures per device.
  6157. - Only display values in the log if they're supported and standardise device log
  6158. line printing.
  6159. Version 2.0.0 - September 6, 2011
  6160. Major feature upgrade - GPU monitoring, (over)clocking and fan control for ATI
  6161. GPUs.
  6162. New command line switches:
  6163. --auto-fan- Automatically adjust all GPU fan speeds to maintain a target
  6164. temperature
  6165. --auto-gpu- Automatically adjust all GPU engine clock speeds to maintain
  6166. a target temperature
  6167. --gpu-engine <arg> Set the GPU engine (over)clock in Mhz - one value for all or
  6168. separate by commas for per card.
  6169. --gpu-fan <arg> Set the GPU fan percentage - one value for all or separate
  6170. by commas for per card.
  6171. --gpu-memclock <arg> Set the GPU memory (over)clock in Mhz - one value for all
  6172. or separate by commas for per card.
  6173. --gpu-powertune <arg> Set the GPU powertune percentage - one value for all or
  6174. separate by commas for per card.
  6175. --gpu-vddc <arg> Set the GPU voltage in Volts - one value for all or separate
  6176. by commas for per card.
  6177. --temp-hysteresis <arg> Set how much the temperature can fluctuate outside
  6178. limits when automanaging speeds (default: 3)
  6179. --temp-overheat <arg> Set the overheat temperature when automatically managing
  6180. fan and GPU speeds (default: 85)
  6181. --temp-target <arg> Set the target temperature when automatically managing fan
  6182. and GPU speeds (default: 75)
  6183. - Implement ATI ADL support for GPU parameter monitoring now and setting later
  6184. (temp, fan, clocks etc.).
  6185. - Check for the presence of the ADL header files in ADL_SDK.
  6186. - Import adl_functions.h from amd overdrive ctrl.
  6187. - Implement a setup function that tries to detect GPUs that support the ADL and
  6188. link in the parameters into the gpus struct.
  6189. - Put a summary of monitoring information from the GPU menu.
  6190. - Implement changing memory speed and voltage on the fly.
  6191. - Implement fan speed setting.
  6192. - Minor corrections to set fan speed by percentage.
  6193. - Make sure to read off the value in RPM only.
  6194. - Implement auto fanspeed adjustment to maintain a target temperature and
  6195. fanspeed below 85%, with an overheat check that will speed the fan up to 100%.
  6196. - Add an --auto-fan command line option to allow all GPUs to have autofan
  6197. enabled from startup.
  6198. - Add a gpu autotune option which adjusts GPU speed to maintain a target
  6199. temperature within the bounds of the default GPU speed and any overclocking set.
  6200. - Avoid a dereference if the longpoll thread doesn't exist.
  6201. - Clean up by setting performance profiles and fan settings to startup levels on
  6202. exit.
  6203. - Add a small amount of hysteresis before lowering clock speed.
  6204. - Allow target, overheat and hysteresis temperatures to be set from command
  6205. line.
  6206. - Combine all stats collating into one function to avoid repeating function
  6207. calls on each variable.
  6208. - Add gpu statistics to debugging output via the watchdog thread.
  6209. - Implement menus to change temperature limits.
  6210. - Implement setting the GPU engine clock speed of all devices or each device as
  6211. a comma separated value.
  6212. - Implement setting the GPU memory clock speed of all devices or each device as
  6213. a comma separated value.
  6214. - Implement setting the GPU voltage of all devices or each device as a comma
  6215. separated value.
  6216. - Implement setting the GPU fan speed of all devices or each device as a comma
  6217. separated value.
  6218. - Add support for monitoring powertune setting.
  6219. - Implement changing of powertune value from the GPU change settings menu.
  6220. - Get the value of powertune in get_stats.
  6221. - Implement setting the GPU powertune value of all devices or each device as a
  6222. comma separated value.
  6223. - Remove the safety checks in speed setting since confirmation is done first in
  6224. the menu, then show the new current values after a short pause.
  6225. - Force the speed to high on startup and restore it to whatever the setting was
  6226. on exit.
  6227. - Add temperature to standard output where possible and use more compact output.
  6228. - Move and print at the same time in curses to avoid random trampling display
  6229. errors.
  6230. - Update the status window only from the watchdog thread, do not rewrite the top
  6231. status messages and only refresh once all the status window is complete,
  6232. clearing the window each time to avoid corruption.
  6233. - Set a safe starting fan speed if we're automanaging the speeds.
  6234. - Provide locking around all adl calls to prevent races.
  6235. - Lower profile settings cannot be higher than higher profile ones so link any
  6236. drops in settings.
  6237. - Add new needed text files to distribution.
  6238. - Queue requests ignoring the number of staged clones since they get discarded
  6239. very easily leading to false positives for pool not providing work fast enough.
  6240. - Include libgen.h in opt.c to fix win32 compilation warnings.
  6241. - Fix compilation warning on win32.
  6242. - Add the directory name from the arguments cgminer was called from as well to
  6243. allow it running from a relative pathname.
  6244. - Add a --disable-adl option to configure and only enable it if opencl support
  6245. exists.
  6246. - Retry before returning a failure to get upstream work as a failure to avoid
  6247. false positives for pool dead.
  6248. - Retry also if the decoding of work fails.
  6249. - Use the presence of X-Roll-Ntime in the header as a bool for exists unless N
  6250. is found in the response.
  6251. Version 1.6.2 - September 2, 2011
  6252. - Add --failover-only option to not leak work to backup pools when the primary
  6253. pool is lagging.
  6254. - Change recommendation to intensity 9 for dedicated miners.
  6255. - Fix the bouncing short term value by allowing it to change dynamically when
  6256. the latest value is very different from the rolling value, but damp the change
  6257. when it gets close.
  6258. - Use the curses_lock to protect the curses_active variable and test it under
  6259. lock.
  6260. - Go back to requesting work 2/3 of the way through the current scantime with
  6261. CPU mining as reports of mining threads running out of work have occurred with
  6262. only 5 seconds to retrieve work.
  6263. - Add start and stop time scheduling for regular time of day running or once off
  6264. start/stop options.
  6265. - Print summary on quit modes.
  6266. - Put some sanity checks on the times that can be input.
  6267. - Give a verbose message when no active pools are found and pause before
  6268. exiting.
  6269. - Add verbose message when a GPU fails to initialise, and disable the correct
  6270. GPU.
  6271. - Cryptopp asm32 was not correctly updated to the incremental nonce code so the
  6272. hash counter was bogus.
  6273. - Get rid of poorly executed curl check.
  6274. - If curl does not have sockopts, do not try to compile the
  6275. json_rpc_call_sockopt_cb function, making it possible to build against older
  6276. curl libraries.
  6277. - Most people expect /usr/local when an unspecified prefix is used so change to
  6278. that.
  6279. - Rename localgen occasions to getwork fail occasions since localgen is
  6280. unrelated now.
  6281. Version 1.6.1 - August 29, 2011
  6282. - Copy cgminer path, not cat it.
  6283. - Switching between redrawing windows does not fix the crash with old
  6284. libncurses, so redraw both windows, but only when the window size hasn't
  6285. changed.
  6286. - Reinstate minimum 1 extra in queue to make it extremely unlikely to ever have
  6287. 0 staged work items and any idle time.
  6288. - Return -1 if no input is detected from the menu to prevent it being
  6289. interpreted as a 0.
  6290. - Make pthread, libcurl and libcurses library checks mandatory or fail.
  6291. - Add a --disable-opencl configure option to make it possible to override
  6292. detection of opencl and build without GPU mining support.
  6293. - Confusion over the variable name for number of devices was passing a bogus
  6294. value which likely was causing the zero sized binary issue.
  6295. - cgminer no longer supports default url user and pass so remove them.
  6296. - Don't show value of intensity since it's dynamic by default.
  6297. - Add options to explicitly enable CPU mining or disable GPU mining.
  6298. - Convert the opt queue into a minimum number of work items to have queued
  6299. instead of an extra number to decrease risk of getting idle devices without
  6300. increasing risk of higher rejects.
  6301. - Statify tv_sort.
  6302. - Check for SSE2 before trying to build 32 bit SSE2 assembly version. Prevents
  6303. build failure when yasm is installed but -msse2 is not specified.
  6304. - Add some defines to configure.ac to enable exporting of values and packaging,
  6305. and clean up output.
  6306. - Give convenient summary at end of ./configure.
  6307. - Display version information and add --version command line option, and make
  6308. sure we flush stdout.
  6309. - Enable curses after the mining threads are set up so that failure messages
  6310. won't be lost in the curses interface.
  6311. - Disable curses after inputting a pool if we requested no curses interface.
  6312. - Add an option to break out after successfully mining a number of accepted
  6313. shares.
  6314. - Exit with a failed return code if we did not reach opt_shares.
  6315. - The cpu mining work data can get modified before we copy it if we submit it
  6316. async, and the sync submission is not truly sync anyway, so just submit it sync.
  6317. Version 1.6.0 - August 26, 2011
  6318. - Make restarting of GPUs optional for systems that hang on any attempt to
  6319. restart them. Fix DEAD status by comparing it to last live time rather than
  6320. last attempted restart time since that happens every minute.
  6321. - Move staged threads to hashes so we can sort them by time.
  6322. - Create a hash list of all the blocks created and search them to detect when a
  6323. new block has definitely appeared, using that information to detect stale work
  6324. and discard it.
  6325. - Update configure.ac for newer autoconf tools.
  6326. - Use the new hashes directly for counts instead of the fragile counters
  6327. currently in use.
  6328. - Update to latest sse2 code from cpuminer-ng.
  6329. - Allow LP to reset block detect and block detect lp flags to know who really
  6330. came first.
  6331. - Get start times just before mining begins to not have very slow rise in
  6332. average.
  6333. - Add message about needing one server.
  6334. - We can queue all the necessary work without hitting frequent stales now with
  6335. the time and string stale protection active all the time. This prevents a
  6336. pool being falsely labelled as not providing work fast enough.
  6337. - Include uthash.h in distro.
  6338. - Implement SSE2 32 bit assembly algorithm as well.
  6339. - Fail gracefully if unable to open the opencl files.
  6340. - Make cgminer look in the install directory for the .cl files making make
  6341. install work correctly.
  6342. - Allow a custom kernel path to be entered on the command line.
  6343. - Bump threshhold for lag up to maximum queued but no staged work.
  6344. - Remove fragile source patching for bitalign, vectors et. al and simply pass it
  6345. with the compiler options.
  6346. - Actually check the value returned for the x-roll-ntime extension to make sure
  6347. it isn't saying N.
  6348. - Prevent segfault on exit for when accessory threads don't exist.
  6349. - Disable curl debugging with opt protocol since it spews to stderr.
  6350. Version 1.5.8 - August 23, 2011
  6351. - Minimise how much more work can be given in cpu mining threads each interval.
  6352. - Make the fail-pause progressively longer each time it fails until the network
  6353. recovers.
  6354. - Only display the lagging message if we've requested the work earlier.
  6355. - Clean up the pool switching to not be dependent on whether the work can roll
  6356. or not by setting a lagging flag and then the idle flag.
  6357. - Only use one thread to determine if a GPU is sick or well, and make sure to
  6358. reset the sick restart attempt time.
  6359. - The worksize was unintentionally changed back to 4k by mistake, this caused a
  6360. slowdown.
  6361. Version 1.5.7 - August 22, 2011
  6362. - Fix a crash with --algo auto
  6363. - Test at appropriate target difficulty now.
  6364. - Add per-device statics log output with --per-device-stats
  6365. - Fix breakage that occurs when 1 or 4 vectors are chosen on new phatk.
  6366. - Make rolltime report debug level only now since we check it every work
  6367. item.
  6368. - Add the ability to enable/disable per-device stats on the fly and match
  6369. logging on/off.
  6370. - Explicitly tell the compiler to retain the program to minimise the chance of
  6371. the zero sized binary errors.
  6372. - Add one more instruction to avoid one branch point in the common path in the
  6373. cl return code. Although this adds more ALUs overall and more branch points, the
  6374. common path code has the same number of ALUs and one less jmp, jmps being more
  6375. expensive.
  6376. - Explicitly link in ws2_32 on the windows build and update README file on how
  6377. to compile successfully on windows.
  6378. - Release cl resources should the gpu mining thread abort.
  6379. - Attempt to restart a GPU once every minute while it's sick.
  6380. - Don't kill off the reinit thread if it fails to init a GPU but returns safely.
  6381. - Only declare a GPU dead if there's been no sign of activity from the reinit
  6382. thread for 10 mins.
  6383. - Never automatically disable any pools but just specify them as idle if they're
  6384. unresponsive at startup.
  6385. - Use any longpoll available, and don't disable it if switching to a server that
  6386. doesn't have it. This allows you to mine solo, yet use the longpoll from a pool
  6387. even if the pool is the backup server.
  6388. - Display which longpoll failed and don't free the ram for lp_url since it
  6389. belongs to the pool hdr path.
  6390. - Make the tcp setsockopts unique to linux in the hope it allows freebsd et. al
  6391. to compile.
  6392. Version 1.5.6 - August 17, 2011
  6393. - New phatk and poclbm kernels. Updated phatk to be in sync with latest 2.2
  6394. courtesy of phateus. Custom modified to work best with cgminer.
  6395. - Updated output buffer code to use a smaller buffer with the kernels.
  6396. - Clean up the longpoll management to ensure the right paths go to the right
  6397. pool and display whether we're connected to LP or not in the status line.
  6398. Version 1.5.5 - August 16, 2011
  6399. - Rework entirely the GPU restart code. Strike a balance between code that
  6400. re-initialises the GPU entirely so that soft hangs in the code are properly
  6401. managed, but if a GPU is completely hung, the thread restart code fails
  6402. gracefully, so that it does not take out any other code or devices. This will
  6403. allow cgminer to keep restarting GPUs that can be restarted, but continue
  6404. mining even if one or more GPUs hangs which would normally require a reboot.
  6405. - Add --submit-stale option which submits all shares, regardless of whether they
  6406. would normally be considered stale.
  6407. - Keep options in alphabetical order.
  6408. - Probe for slightly longer for when network conditions are lagging.
  6409. - Only display the CPU algo when we're CPU mining.
  6410. - As we have keepalives now, blaming network flakiness on timeouts appears to
  6411. have been wrong. Set a timeout for longpoll to 1 hour, and most other
  6412. network connectivity to 1 minute.
  6413. - Simplify output code and remove HW errors from CPU stats.
  6414. - Simplify code and tidy output.
  6415. - Only show cpu algo in summary if cpu mining.
  6416. - Log summary at the end as per any other output.
  6417. - Flush output.
  6418. - Add a linux-usb-cgminer guide courtesy of Kano.
  6419. Version 1.5.4 - August 14, 2011
  6420. - Add new option: --monitor <cmd> Option lets user specify a command <cmd> that
  6421. will get forked by cgminer on startup. cgminer's stderr output subsequently gets
  6422. piped directly to this command.
  6423. - Allocate work from one function to be able to initialise variables added
  6424. later.
  6425. - Add missing fflush(stdout) for --ndevs and conclusion summary.
  6426. - Preinitialise the devices only once on startup.
  6427. - Move the non cl_ variables into the cgpu info struct to allow creating a new
  6428. cl state on reinit, preserving known GPU variables.
  6429. - Create a new context from scratch in initCQ in case something was corrupted to
  6430. maximise our chance of succesfully creating a new worker thread. Hopefully this
  6431. makes thread restart on GPU failure more reliable, without hanging everything
  6432. in the case of a completely wedged GPU.
  6433. - Display last initialised time in gpu management info, to know if a GPU has
  6434. been re-initialised.
  6435. - When pinging a sick cpu, flush finish and then ping it in a separate thread in
  6436. the hope it recovers without needing a restart, but without blocking code
  6437. elsewhere.
  6438. - Only consider a pool lagging if we actually need the work and we have none
  6439. staged despite queue requests stacking up. This decreases significantly the
  6440. amount of work that leaks to the backup pools.
  6441. - The can_roll function fails inappropriately in stale_work.
  6442. - Only put the message that a pool is down if not pinging it every minute. This
  6443. prevents cgminer from saying pool down at 1 minute intervals unless in debug
  6444. mode.
  6445. - Free all work in one place allowing us to perform actions on it in the future.
  6446. - Remove the extra shift in the output code which was of dubious benefit. In
  6447. fact in cgminer's implementation, removing this caused a miniscule speedup.
  6448. - Test each work item to see if it can be rolled instead of per-pool and roll
  6449. whenever possible, adhering to the 60 second timeout. This makes the period
  6450. after a longpoll have smaller dips in throughput, as well as requiring less
  6451. getworks overall thus increasing efficiency.
  6452. - Stick to rolling only work from the current pool unless we're in load balance
  6453. mode or lagging to avoid aggressive rolling imitating load balancing.
  6454. - If a work item has had any mining done on it, don't consider it discarded
  6455. work.
  6456. Version 1.5.3 - July 30, 2011
  6457. - Significant work went into attempting to make the thread restart code robust
  6458. to identify sick threads, tag them SICK after 1 minute, then DEAD after 5
  6459. minutes of inactivity and try to restart them. Instead of re-initialising the
  6460. GPU completely, only a new cl context is created to avoid hanging the rest of
  6461. the GPUs should the dead GPU be hung irrevocably.
  6462. - Use correct application name in syslog.
  6463. - Get rid of extra line feeds.
  6464. - Use pkg-config to check for libcurl version
  6465. - Implement per-thread getwork count with proper accounting to not over-account
  6466. queued items when local work replaces it.
  6467. - Create a command queue from the program created from source which allows us
  6468. to flush the command queue in the hope it will not generate a zero sized binary
  6469. any more.
  6470. - Be more willing to get work from the backup pools if the work is simply being
  6471. queued faster than it is being retrieved.
  6472. Version 1.5.2 - July 28, 2011
  6473. - Restarting a hung GPU can hang the rest of the GPUs so just declare it dead
  6474. and provide the information in the status.
  6475. - The work length in the miner thread gets smaller but doesn't get bigger if
  6476. it's under 1 second. This could end up leading to CPU under-utilisation and
  6477. lower and lower hash rates. Fix it by increasing work length if it drops
  6478. under 1 second.
  6479. - Make the "quiet" mode still update the status and display errors, and add a
  6480. new --real-quiet option which disables all output and can be set once while
  6481. running.
  6482. - Update utility and efficiency figures when displaying them.
  6483. - Some Intel HD graphics support the opencl commands but return errors since
  6484. they don't support opencl. Don't fail with them, just provide a warning and
  6485. disable GPU mining.
  6486. - Add http:// if it's not explicitly set for URL entries.
  6487. - Log to the output file at any time with warnings and errors, instead of just
  6488. when verbose mode is on.
  6489. - Display the correct current hash as per blockexplorer, truncated to 16
  6490. characters, with just the time.
  6491. Version 1.5.1 - July 27, 2011
  6492. - Two redraws in a row cause a crash in old libncurses so just do one redraw
  6493. using the main window.
  6494. - Don't adjust hash_div only up for GPUs. Disable hash_div adjustment for GPUs.
  6495. - Only free the thread structures if the thread still exists.
  6496. - Update both windows separately, but not at the same time to prevent the double
  6497. refresh crash that old libncurses has. Do the window resize check only when
  6498. about to redraw the log window to minimise ncurses cpu usage.
  6499. - Abstract out the decay time function and use it to make hash_div a rolling
  6500. average so it doesn't change too abruptly and divide work in chunks large enough
  6501. to guarantee they won't overlap.
  6502. - Sanity check to prove locking.
  6503. - Don't take more than one lock at a time.
  6504. - Make threads report out when they're queueing a request and report if they've
  6505. failed.
  6506. - Make cpu mining work submission asynchronous as well.
  6507. - Properly detect stale work based on time from staging and discard instead of
  6508. handing on, but be more lax about how long work can be divided for up to the
  6509. scantime.
  6510. - Do away with queueing work separately at the start and let each thread grab
  6511. its own work as soon as it's ready.
  6512. - Don't put an extra work item in the queue as each new device thread will do so
  6513. itself.
  6514. - Make sure to decrease queued count if we discard the work.
  6515. - Attribute split work as local work generation.
  6516. - If work has been cloned it is already at the head of the list and when being
  6517. reinserted into the queue it should be placed back at the head of the list.
  6518. - Dividing work is like the work is never removed at all so treat it as such.
  6519. However the queued bool needs to be reset to ensure we *can* request more work
  6520. even if we didn't initially.
  6521. - Make the display options clearer.
  6522. - Add debugging output to tq_push calls.
  6523. - Add debugging output to all tq_pop calls.
  6524. Version 1.5.0 - July 26, 2011
  6525. - Increase efficiency of slow mining threads such as CPU miners dramatically. Do
  6526. this by detecting which threads cannot complete searching a work item within the
  6527. scantime and then divide up a work item into multiple smaller work items.
  6528. Detect the age of the work items and if they've been cloned before to prevent
  6529. doing the same work over. If the work is too old to be divided, then see if it
  6530. can be time rolled and do that to generate work. This dramatically decreases the
  6531. number of queued work items from a pool leading to higher overall efficiency
  6532. (but the same hashrate and share submission rate).
  6533. - Don't request work too early for CPUs as CPUs will scan for the full
  6534. opt_scantime anyway.
  6535. - Simplify gpu management enable/disable/restart code.
  6536. - Implement much more accurate rolling statistics per thread and per gpu and
  6537. improve accuracy of rolling displayed values.
  6538. - Make the rolling log-second average more accurate.
  6539. - Add a menu to manage GPUs on the fly allowing you to enable/disable GPUs or
  6540. try restarting them.
  6541. - Keep track of which GPUs are alive versus enabled.
  6542. - Start threads for devices that are even disabled, but don't allow them to
  6543. start working.
  6544. - The last pool is when we are low in total_pools, not active_pools.
  6545. - Make the thread restart do a pthread_join after disabling the device, only
  6546. re-enabling it if we succeed in restarting the thread. Do this from a separate
  6547. thread so as to not block any other code.This will allow cgminer to continue
  6548. even if one GPU hangs.
  6549. - Try to do every curses manipulation under the curses lock.
  6550. - Only use the sockoptfunction if the version of curl is recent enough.
  6551. Version 1.4.1 - July 24, 2011
  6552. - Do away with GET for dealing with longpoll forever. POST is the one that works
  6553. everywhere, not the other way around.
  6554. - Detect when the primary pool is lagging and start queueing requests on backup
  6555. pools if possible before needing to roll work.
  6556. - Load balancing puts more into the current pool if there are disabled pools.
  6557. Fix.
  6558. - Disable a GPU device should the thread fail to init.
  6559. - Out of order command queue may fail on osx. Try without if it fails.
  6560. - Fix possible dereference on blank inputs during input_pool.
  6561. - Defines missing would segfault on --help when no sse mining is built in.
  6562. - Revert "Free up resources/stale compilers." - didn't help.
  6563. - Only try to print the status of active devices or it would crash.
  6564. - Some hardware might benefit from the less OPS so there's no harm in leaving
  6565. kernel changes that do that apart from readability of the code.
  6566. Version 1.4.0 - July 23, 2011
  6567. - Feature upgrade: Add keyboard input during runtime to allow modification of
  6568. and viewing of numerous settings such as adding/removing pools, changing
  6569. multipool management strategy, switching pools, changing intensiy, verbosity,
  6570. etc. with a simple keypress menu system.
  6571. - Free up resources/stale compilers.
  6572. - Kernels are safely flushed in a way that allows out of order execution to
  6573. work.
  6574. - Sometimes the cl compiler generates zero sized binaries and only a reboot
  6575. seems to fix it.
  6576. - Don't try to stop/cancel threads that don't exist.
  6577. - Only set option to show devices and exit if built with opencl support.
  6578. - Enable curses earlier and exit with message in main for messages to not be
  6579. lost in curses windows.
  6580. - Make it possible to enter server credentials with curses input if none are
  6581. specified on the command line.
  6582. - Abstract out a curses input function and separate input pool function to allow
  6583. for live adding of pools later.
  6584. - Remove the nil arguments check to allow starting without parameters.
  6585. - Disable/enable echo & cbreak modes.
  6586. - Add a thread that takes keyboard input and allow for quit, silent, debug,
  6587. verbose, normal, rpc protocol debugging and clear screen options.
  6588. - Add pool option to input and display current pool status, pending code to
  6589. allow live changes.
  6590. - Add a bool for explicit enabling/disabling of pools.
  6591. - Make input pool capable of bringing up pools while running.
  6592. - Do one last check of the work before submitting it.
  6593. - Implement the ability to live add, enable, disable, and switch to pools.
  6594. - Only internally test for block changes when the work matches the current pool
  6595. to prevent interleaved block change timing on multipools.
  6596. - Display current pool management strategy to enable changing it on the fly.
  6597. - The longpoll blanking of the current_block data may not be happening before
  6598. the work is converted and appears to be a detected block change. Blank the
  6599. current block be
  6600. - Make --no-longpoll work again.
  6601. - Abstract out active pools count.
  6602. - Allow the pool strategy to be modified on the fly.
  6603. - Display pool information on the fly as well.
  6604. - Add a menu and separate out display options.
  6605. - Clean up the messy way the staging thread communicates with the longpoll
  6606. thread to determine who found the block first.
  6607. - Make the input windows update immediately instead of needing a refresh.
  6608. - Allow log interval to be set in the menu.
  6609. - Allow scan settings to be modified at runtime.
  6610. - Abstract out the longpoll start and explicitly restart it on pool change.
  6611. - Make it possible to enable/disable longpoll.
  6612. - Set priority correctly on multipools. Display priority and alive/dead
  6613. information in display_pools.
  6614. - Implement pool removal.
  6615. - Limit rolltime work generation to 10 iterations only.
  6616. - Decrease testing log to info level.
  6617. - Extra refresh not required.
  6618. - With huge variation in GPU performance, allow intensity to go from -10 to +10.
  6619. - Tell getwork how much of a work item we're likely to complete for future
  6620. splitting up of work.
  6621. - Remove the mandatory work requirement at startup by testing for invalid work
  6622. being passed which allows for work to be queued immediately. This also
  6623. removes the requirem
  6624. - Make sure intensity is carried over to thread count and is at least the
  6625. minimum necessary to work.
  6626. - Unlocking error on retry. Locking unnecessary anyway so remove it.
  6627. - Clear log window from consistent place. No need for locking since logging is
  6628. disabled during input.
  6629. - Cannot print the status of threads that don't exist so just queue enough work
  6630. for the number of mining threads to prevent crash with -Q N.
  6631. - Update phatk kernel to one with new parameters for slightly less overhead
  6632. again. Make the queue kernel parameters call a function pointer to select
  6633. phatk or poclbm.
  6634. - Make it possible to select the choice of kernel on the command line.
  6635. - Simplify the output part of the kernel. There's no demonstrable advantage from
  6636. more complexity.
  6637. - Merge pull request #18 from ycros/cgminer
  6638. - No need to make leaveok changes win32 only.
  6639. - Build support in for all SSE if possible and only set the default according to
  6640. machine capabilities.
  6641. - Win32 threading and longpoll keepalive fixes.
  6642. - Win32: Fix for mangled output on the terminal on exit.
  6643. Version 1.3.1 - July 20, 2011
  6644. - Feature upgrade; Multiple strategies for failover. Choose from default which
  6645. now falls back to a priority order from 1st to last, round robin which only
  6646. changes pools when one is idle, rotate which changes pools at user-defined
  6647. intervals, and load-balance which spreads the work evenly amongst all pools.
  6648. - Implement pool rotation strategy.
  6649. - Implement load balancing algorithm by rotating requests to each pool.
  6650. - Timeout on failed discarding of staged requests.
  6651. - Implement proper flagging of idle pools, test them with the watchdog thread,
  6652. and failover correctly.
  6653. - Move pool active test to own function.
  6654. - Allow multiple strategies to be set for multipool management.
  6655. - Track pool number.
  6656. - Don't waste the work items queued on testing the pools at startup.
  6657. - Reinstate the mining thread watchdog restart.
  6658. - Add a getpoll bool into the thread information and don't restart threads stuck
  6659. waiting on work.
  6660. - Rename the idlenet bool for the pool for later use.
  6661. - Allow the user/pass userpass urls to be input in any order.
  6662. - When json rpc errors occur they occur in spits and starts, so trying to limit
  6663. them with the comms error bool doesn't stop a flood of them appearing.
  6664. - Reset the queued count to allow more work to be queued for the new pool on
  6665. pool switch.
  6666. Version 1.3.0 - July 19, 2011
  6667. - Massive infrastructure update to support pool failover.
  6668. - Accept multiple parameters for url, user and pass and set up structures of
  6669. pool data accordingly.
  6670. - Probe each pool for what it supports.
  6671. - Implement per pool feature support according to rolltime support as
  6672. advertised by server.
  6673. - Do switching automatically based on a 300 second timeout of locally generated
  6674. work or 60 seconds of no response from a server that doesn't support rolltime.
  6675. - Implement longpoll server switching.
  6676. - Keep per-pool data and display accordingly.
  6677. - Make sure cgminer knows how long the pool has actually been out for before
  6678. deeming it a prolonged outage.
  6679. - Fix bug with ever increasing staged work in 1.2.8 that eventually caused
  6680. infinite rejects.
  6681. - Make warning about empty http requests not show by default since many
  6682. servers do this regularly.
  6683. Version 1.2.8 - July 18, 2011
  6684. - More OSX build fixes.
  6685. - Add an sse4 algorithm to CPU mining.
  6686. - Fix CPU mining with other algorithms not working.
  6687. - Rename the poclbm file to ensure a new binary is built since.
  6688. - We now are guaranteed to have one fresh work item after a block change and we
  6689. should only discard staged requests.
  6690. - Don't waste the work we retrieve from a longpoll.
  6691. - Provide a control lock around global bools to avoid racing on them.
  6692. - Iterating over 1026 nonces when confirming data from the GPU is old code
  6693. and unnecessary and can lead to repeats/stales.
  6694. - The poclbm kernel needs to be updated to work with the change to 4k sized
  6695. output buffers.
  6696. - longpoll seems to work either way with post or get but some servers prefer
  6697. get so change to httpget.
  6698. Version 1.2.7 - July 16, 2011
  6699. - Show last 8 characters of share submitted in log.
  6700. - Display URL connected to and user logged in as in status.
  6701. - Display current block and when it was started in the status line.
  6702. - Only pthread_join the mining threads if they exist as determined by
  6703. pthread_cancel and don't fail on pthread_cancel.
  6704. - Create a unique work queue for all getworks instead of binding it to thread 0
  6705. to avoid any conflict over thread 0's queue.
  6706. - Clean up the code to make it clear it's watchdog thread being messaged to
  6707. restart the threads.
  6708. - Check the current block description hasn't been blanked pending the real
  6709. new current block data.
  6710. - Re-enable signal handlers once the signal has been received to make it
  6711. possible to kill cgminer if it fails to shut down.
  6712. - Disable restarting of CPU mining threads pending further investigation.
  6713. - Update longpoll messages.
  6714. - Add new block data to status line.
  6715. - Fix opencl tests for osx.
  6716. - Only do local generation of work if the work item is not stale itself.
  6717. - Check for stale work within the mining threads and grab new work if
  6718. positive.
  6719. - Test for idle network conditions and prevent threads from being restarted
  6720. by the watchdog thread under those circumstances.
  6721. - Make sure that local work generation does not continue indefinitely by
  6722. stopping it after 10 minutes.
  6723. - Tweak the kernel to have a shorter path using a 4k buffer and a mask on the
  6724. nonce value instead of a compare and loop for a shorter code path.
  6725. - Allow queue of zero and make that default again now that we can track how
  6726. work is being queued versus staged. This can decrease reject rates.
  6727. - Queue precisely the number of mining threads as longpoll_staged after a
  6728. new block to not generate local work.
  6729. Version 1.2.6 - July 15, 2011
  6730. - Put a current system status line beneath the total work status line
  6731. - Fix a counting error that would prevent cgminer from correctly detecting
  6732. situations where getwork was failing - this would cause stalls sometimes
  6733. unrecoverably.
  6734. - Limit the maximum number of requests that can be put into the queue which
  6735. otherwise could get arbitrarily long during a network outage.
  6736. - Only count getworks that are real queue requests.
  6737. Version 1.2.5 - July 15, 2011
  6738. - Conflicting -n options corrected
  6739. - Setting an intensity with -I disables dynamic intensity setting
  6740. - Removed option to manually disable dynamic intensity
  6741. - Improve display output
  6742. - Implement signal handler and attempt to clean up properly on exit
  6743. - Only restart threads that are not stuck waiting on mandatory getworks
  6744. - Compatibility changes courtesy of Ycros to build on mingw32 and osx
  6745. - Explicitly grab first work item to prevent false positive hardware errors
  6746. due to working on uninitialised work structs
  6747. - Add option for non curses --text-only output
  6748. - Ensure we connect at least once successfully before continuing to retry to
  6749. connect in case url/login parameters were wrong
  6750. - Print an executive summary when cgminer is terminated
  6751. - Make sure to refresh the status window
  6752. Versions -> 1.2.4
  6753. - Con Kolivas - July 2011. New maintainership of code under cgminer name.
  6754. - Massive rewrite to incorporate GPU mining.
  6755. - Incorporate original oclminer c code.
  6756. - Rewrite gpu mining code to efficient work loops.
  6757. - Implement per-card detection and settings.
  6758. - Implement vector code.
  6759. - Implement bfi int patching.
  6760. - Import poclbm and phatk ocl kernels and use according to hardware type.
  6761. - Implement customised optimised versions of opencl kernels.
  6762. - Implement binary kernel generation and loading.
  6763. - Implement preemptive asynchronous threaded work gathering and pushing.
  6764. - Implement variable length extra work queues.
  6765. - Optimise workloads to be efficient miners instead of getting lots of extra
  6766. work.
  6767. - Implement total hash throughput counters, per-card accepted, rejected and
  6768. hw error count.
  6769. - Staging and watchdog threads to prevent fallover.
  6770. - Stale and reject share guarding.
  6771. - Autodetection of new blocks without longpoll.
  6772. - Dynamic setting of intensity to maintain desktop interactivity.
  6773. - Curses interface with generous statistics and information.
  6774. - Local generation of work (xroll ntime) when detecting poor network
  6775. connectivity.
  6776. Version 1.0.2
  6777. - Linux x86_64 optimisations - Con Kolivas
  6778. - Optimise for x86_64 by default by using sse2_64 algo
  6779. - Detects CPUs and sets number of threads accordingly
  6780. - Uses CPU affinity for each thread where appropriate
  6781. - Sets scheduling policy to lowest possible
  6782. - Minor performance tweaks
  6783. Version 1.0.1 - May 14, 2011
  6784. - OSX support
  6785. Version 1.0 - May 9, 2011
  6786. - jansson 2.0 compatibility
  6787. - correct off-by-one in date (month) display output
  6788. - fix platform detection
  6789. - improve yasm configure bits
  6790. - support full URL, in X-Long-Polling header
  6791. Version 0.8.1 - March 22, 2011
  6792. - Make --user, --pass actually work
  6793. - Add User-Agent HTTP header to requests, so that server operators may
  6794. more easily identify the miner client.
  6795. - Fix minor bug in example JSON config file
  6796. Version 0.8 - March 21, 2011
  6797. - Support long polling: http://deepbit.net/longpolling.php
  6798. - Adjust max workload based on scantime (default 5 seconds,
  6799. or 60 seconds for longpoll)
  6800. - Standardize program output, and support syslog on Unix platforms
  6801. - Suport --user/--pass options (and "user" and "pass" in config file),
  6802. as an alternative to the current --userpass
  6803. Version 0.7.2 - March 14, 2011
  6804. - Add port of ufasoft's sse2 assembly implementation (Linux only)
  6805. This is a substantial speed improvement on Intel CPUs.
  6806. - Move all JSON-RPC I/O to separate thread. This reduces the
  6807. number of HTTP connections from one-per-thread to one, reducing resource
  6808. usage on upstream bitcoind / pool server.
  6809. Version 0.7.1 - March 2, 2011
  6810. - Add support for JSON-format configuration file. See example
  6811. file example-cfg.json. Any long argument on the command line
  6812. may be stored in the config file.
  6813. - Timestamp each solution found
  6814. - Improve sha256_4way performance. NOTE: This optimization makes
  6815. the 'hash' debug-print output for sha256_way incorrect.
  6816. - Use __builtin_expect() intrinsic as compiler micro-optimization
  6817. - Build on Intel compiler
  6818. - HTTP library now follows HTTP redirects
  6819. Version 0.7 - February 12, 2011
  6820. - Re-use CURL object, thereby reuseing DNS cache and HTTP connections
  6821. - Use bswap_32, if compiler intrinsic is not available
  6822. - Disable full target validation (as opposed to simply H==0) for now
  6823. Version 0.6.1 - February 4, 2011
  6824. - Fully validate "hash < target", rather than simply stopping our scan
  6825. if the high 32 bits are 00000000.
  6826. - Add --retry-pause, to set length of pause time between failure retries
  6827. - Display proof-of-work hash and target, if -D (debug mode) enabled
  6828. - Fix max-nonce auto-adjustment to actually work. This means if your
  6829. scan takes longer than 5 seconds (--scantime), the miner will slowly
  6830. reduce the number of hashes you work on, before fetching a new work unit.
  6831. Version 0.6 - January 29, 2011
  6832. - Fetch new work unit, if scanhash takes longer than 5 seconds (--scantime)
  6833. - BeeCee1's sha256 4way optimizations
  6834. - lfm's byte swap optimization (improves via, cryptopp)
  6835. - Fix non-working short options -q, -r
  6836. Version 0.5 - December 28, 2010
  6837. - Exit program, when all threads have exited
  6838. - Improve JSON-RPC failure diagnostics and resilience
  6839. - Add --quiet option, to disable hashmeter output.
  6840. Version 0.3.3 - December 27, 2010
  6841. - Critical fix for sha256_cryptopp 'cryptopp_asm' algo
  6842. Version 0.3.2 - December 23, 2010
  6843. - Critical fix for sha256_via
  6844. Version 0.3.1 - December 19, 2010
  6845. - Critical fix for sha256_via
  6846. - Retry JSON-RPC failures (see --retry, under "minerd --help" output)
  6847. Version 0.3 - December 18, 2010
  6848. - Add crypto++ 32bit assembly implementation
  6849. - show version upon 'minerd --help'
  6850. - work around gcc 4.5.x bug that killed 4way performance
  6851. Version 0.2.2 - December 6, 2010
  6852. - VIA padlock implementation works now
  6853. - Minor build and runtime fixes
  6854. Version 0.2.1 - November 29, 2010
  6855. - avoid buffer overflow when submitting solutions
  6856. - add Crypto++ sha256 implementation (C only, ASM elided for now)
  6857. - minor internal optimizations and cleanups
  6858. Version 0.2 - November 27, 2010
  6859. - Add script for building a Windows installer
  6860. - improve hash performance (hashmeter) statistics
  6861. - add tcatm 4way sha256 implementation
  6862. - Add experimental VIA Padlock sha256 implementation
  6863. Version 0.1.2 - November 26, 2010
  6864. - many small cleanups and micro-optimizations
  6865. - build win32 exe using mingw
  6866. - RPC URL, username/password become command line arguments
  6867. - remove unused OpenSSL dependency
  6868. Version 0.1.1 - November 24, 2010
  6869. - Do not build sha256_generic module separately from cpuminer.
  6870. Version 0.1 - November 24, 2010
  6871. - Initial release.