Gbeɖiɖi Kple Nya Siwo Ŋu Dɔwɔna Siwo Woatsɔ Adzro C Dɔwɔɖoɖo Siwo Me De blibo O Ŋu [pdf].

| Abe alesi wòle be static analysis tools nabu tame le nuwɔna siwo womeɖe fia o kple modules siwo bu le C codebases me ŋu ene la, egbegbe asitsalawo hã dzea ŋgɔ kuxi vevi ma ke — dɔwɔwɔ kple ɖoɖo siwo "mewu enu" bliboe gbeɖe o la wɔwɔ kple wo nyonyo.

Nukae Nye Points-To Analysis Kple Nukatae Wòle Vevie Na Egbegbe Dɔwɔwɔwo?

Points-to analysis nye static program analysis ƒomevi aɖe si dea dzesi ŋkuɖodzi teƒe siwo pointer variable ateŋu ayɔ le ewɔwɔɣi. Le C ɖoɖowɔɖi siwo mede blibo o ƒe nyawo me — bu agbalẽdzraɖoƒewo, codebases ƒe akpa aɖe, alo ɖoɖo siwo ŋu nusiwo dzi woanɔ te ɖo bu gbɔ — "gbeɖiɖi" numekuku gbɔ ɖoɖo fia be màto nufiamɔ̃ ƒe ƒomedodo si sɔ ŋu gbeɖe o, ne ema fia be woatsɔe asɔ kple wo nɔewo wògbɔ eme ɣeaɖewoɣi gɔ̃ hã. Gbeɖiɖi ƒe nukpɔsusu le vevie ŋutɔ: gbeɖiɖi me dzodzro meɖea alakpa nya manyomanyo siwo ate ŋu atsyɔ kuxi ŋutɔŋutɔwo dzi gbeɖe o.

| Sukugbalẽwo, siwo dome gɔmeɖoanyi PDF gbalẽ siwo tso dɔwɔƒewo abe Carnegie Mellon kple ETH Zurich gbɔ hã le, ɖee fia be gbeɖiɖi kple scalability siaa gbɔ ɖoɖo bia abstractions siwo ŋu wotrɔ asi le nyuie — vevietɔ ƒo xlã heap modeling, context sensitivity, kple unknown function handling.

ƒe nyawo

"Kukuɖenuŋu si me kɔ akpa be woawɔ nu ɖe eŋu la menye numekuku kura o. Taɖodzinua menye nyateƒetoto le nukpɔsusu me ko o, ke boŋ viɖe ŋutɔŋutɔ le xexeame ŋutɔŋutɔ ƒe mɔxenuwo te — eɖanye be èle C ƒe fli miliɔn geɖe me dzrom alo nèle asitsadɔ si le dzidzim ɖe edzi si me nyatakaka siwo mede blibo o le dzi kpɔm o."

ƒe nyawo

Aleke Dɔwɔɖoɖo Siwo Mede blibo O Naa Kakaɖedzimanɔamesi si Le Cascading Le Kukuɖenuŋu Me?

Kuxi si nye blibomademade le C ɖoɖowɔɖi ƒe numekuku me ɖe dɔwɔwɔ ƒe nyateƒe si keke ta wu fia: ɖoɖo akpa gãtɔ — kɔmpiuta dɔwɔɖoɖo alo habɔbɔ — mele nɔnɔme si wowu enu me gbeɖe o. Ne static analyzer do go gotagome kaƒoƒo si me kpukpui aɖeke mele o la, ele be wòawɔ susu siwo me kɔ. Susu siawo kaka to numekukua me, si ate ŋu ana points-to sets la nadzi ɖe edzi eye wòaɖe woƒe nyateƒetoto dzi akpɔtɔ. Kakaɖedzimanɔamesi ma dzi kpɔkpɔ evɔ womatsɔ nu nyui wɔwɔ asa vɔe o ye nye mɔ̃ɖaŋudɔwɔwɔ ƒe kuxi vevitɔ.

Aɖaŋu siwo wozã le numekuku gbalẽwo me dometɔ aɖewoe nye:

ƒe nyawo
• Conservative external function modeling — nuwɔwɔ ɖe dɔwɔwɔ siwo womenya o ŋu be woateŋu atrɔ asi le heap teƒe ɖesiaɖe si woateŋu aɖo
• Numekuku si wotu ɖe didi dzi — akɔntabubu si fiaa asi nyatakakawo na biabia siwo le vevie ko, si ɖea gazazã dzi kpɔtɔna le kɔdaɖoɖo siwo mede blibo o dzi
• Context-sensitive heap abstraction — vovototodedeameme le mama teƒewo to yɔyɔ na context be woaɖe alakpa aliasing dzi akpɔtɔ
• Dzidziɖedzi ƒe ɖɔɖɔɖo — dze egɔme kple akɔtadzesi si gbɔ eme si le gbadzaa eye woakɔ eŋu le afisiwo wòhiã be woawɔ nu pɛpɛpɛ le na asisi ƒe numekuku si wona ko
• Stub dzidzi na modules siwo bu — ƒoƒo ƒu kpukpui siwo me kɔ na dzesi siwo womeɖe o le kadodo ƒe liƒowo dzi

ƒe nyawo

Aɖaŋu ɖesiaɖe ɖea asitsatsa si le nyonyome, nuwɔwɔ pɛpɛpɛ, kple akɔntabubu ƒe gazazã dome fiana — asitsatsa si dzena le akpa ɖesiaɖe kloe me afisi wòle be woawɔ nyametsotsowo le kakaɖedzimanɔamesi te.

Nusɔsrɔ̃ Kawoe Asitsahabɔbɔwo Ate Ŋu Akpɔ Tso Gbeɖiɖiwo Me Dzodzro ƒe Gɔmeɖosewo Me?

Amehehe si hiã be woatu gbeɖiɖi ƒe static analyzers la ɖea egɔme tẽ ɖe asitsatsa ƒe dɔwɔna nyuiwo me. Asitsaha si mateŋu "adzro" eya ŋutɔ ƒe nɔnɔme me o — asisiwo, gakpɔkpɔ, ƒuƒoƒo ƒe dɔwɔwɔ, kple adzɔnuwo ƒe xexlẽdzesiwo yometiti — le dɔ wɔm le nyatakaka siwo mede blibo o dzi, abe alesi numekula si doa go dzesi siwo womeɖe o ene. Afɔku si le asitsatsa ŋuti nuŋububu si mesɔ o me la sɔ kple nufiamɔ̃ ƒe numekuku si mesɔ o: ètoa kuxi ŋutɔŋutɔwo ŋu vaseɖe esime wozua kpododonu sesẽwo.

Asitsaha siwo le dzidzim ɖe edzi la le avu wɔm kple dɔwɔnu siwo me mama le ɣesiaɣi — mɔnu ɖeka na CRM, bubu na numekuku, bubu na dɔa dzikpɔkpɔ, bubu na asitsatsa ƒe nuwo wɔwɔ le wo ɖokui si. Vovototo ɖesiaɖe si le ɖekawɔwɔ me nye "module si bu" le dɔwɔwɔ ƒe numekuku me. Nyametsotso siwo wowɔ le nyatakaka ƒe akpa aɖewo ŋu, le gɔmesese nu la, ate ŋu anye esiwo mesɔ o. Nusi woɖɔ ɖo, le ɖoɖowɔɖiwo me dzodzro kple asitsatsa siaa me, enye ƒoƒo ƒu ɖekae: nyatakaka siwo katã sɔ la tsɔtsɔ va kpɔɖeŋu ɖeka si wɔ ɖeka me.

Aleke Asitsatsa ƒe Dɔwɔƒe Ðeka Kpɔa Blibomakpɔmakpɔ ƒe Kuxia Gbɔe?

Wotu Mewayz ɖe gɔmesese sia tututu dzi. Abe asitsadɔwɔɖoɖo si katã le ɖeka me si me modules 207 siwo wotsɔ wɔ ɖekae kple ezãla siwo wu 138,000 siwo le dɔ dzi le ene la, Mewayz ɖea blibomademade ƒe kuxi si ɖea fu na asitsaha siwo le dɔ wɔm le SaaS stack siwo metso kadodo me o la ɖa. Ne wò CRM, nyatakakawo dzikpɔkpɔ, hadomenyatakakadzraɖoƒewo ƒe ɖoɖowɔwɔ, e-asitsatsa, ƒuƒoƒo ƒe nuwɔwɔ aduadu, numekuku, kple asisiwo ƒe nyatakakadzraɖoƒe katã le mɔnu ɖeka me la, ke èle — le asitsatsa ƒe nya nu — ɖoɖowɔɖi bliboa ƒe numekuku wɔm tsɔ wu be nàwɔ module ɖesiaɖe ƒe tsɔtsɔ sɔ kple wo nɔewo.

Emetsonue nye asitsatsa ŋuti nyametsotso nyuiwo wɔwɔ. Ènya asisi siwo le afɔku me be woaʋuʋu wo elabena kpekpeɖeŋu ŋuti nyatakakawo, wo zazã ŋuti nyatakakawo, kple fexexe ŋuti nyatakakawo katã dzena le teƒe ɖeka. Ènya asitsatsa ƒe aʋawɔnya kawoe le gakpɔkpɔ hem ŋutɔŋutɔ elabena kɔsɔkɔsɔ si wotsɔ gblɔa nusiwo gbɔ wòtso la metso o. "Gotagome dɔwɔwɔ yɔyɔ" aɖeke meli si me kpukpui siwo womenya o le — asitsatsa ƒe akpa ɖesiaɖe ɖua nu ɖe kpɔɖeŋu ɖeka si wɔ ɖeka me.

|

Nukae Nye Etsɔme Nɔnɔme Siwo Do Ƒome Kple Kɔmpiutadziɖoɖowo Ŋuti Numekuku Kple Asitsatsa Ŋuti Nunya?

Mɔnu siwo wozãna le se nu kple asitsatsa ƒe dɔwɔnawo ƒe ƒoƒo ɖekae le kabakabam. AI-kpekpeɖeŋunana static analysis dɔwɔnuwo le dɔwɔwɔ ƒe kpukpuiwo wɔwɔ gɔme na code si bu le wo ɖokui si, si ɖe blibomademade ƒe kuxia dzi kpɔtɔ le ɖoɖowɔɖi ƒe numekuku me ŋutɔ. Wole AI ƒe ŋutete mawo ke zãm ɖe asitsatsa ŋuti nunya ŋu — wole nu ƒom tso kpɔɖeŋuwo ŋu, wole vovototo siwo le nyatakakawo me la yɔyɔ, kple gɔmesese siwo dɔwɔnu siwo kakɛkakɛwo ato keŋkeŋ la dodo ɖe ŋgɔ.

Nuƒolanɔƒewo abe Mewayz ene le ŋgɔ na nɔnɔme sia, tsɔa AI-dzɔdzɔme dɔwɔwɔwo ƒoa ƒu tẽe ɖe asitsatsa ƒe OS ɖekawɔwɔ me. Taɖodzinua sɔ kple esi le gbeɖiɖi ƒe nɔnɔme me dzodzro me: nyateƒetoto si ŋu viɖe le wu kple alakpa manyomanyo suetɔ kekeake, ale be wotu asitsatsa ŋuti nyametsotso ɖesiaɖe ɖe nyatakaka blibo siwo ŋu kakaɖedzi le dzi.

Nyabiase Siwo Wobiana Enuenu

Nukae na be points-to analysis "ɖina" na C ɖoɖowɔɖi siwo mede blibo o?

Gbeɖiɖi ƒe dzesi-ɖe-ku-ɖe-nusɔsrɔ̃ naa kakaɖedzi be meto ŋkɔ bubu alo nufiamɔ̃ ƒe ƒomedodo si sɔ ŋu gbeɖe o. Le dɔwɔɖoɖo siwo mede blibo o — esiwo si dzɔtsoƒe faɛlwo bu, gota agbalẽdzraɖoƒewo, alo gegeɖeƒe siwo womeɖe o — gome la, gbeɖiɖi ƒe nyonyome bia be woatsɔ susu siwo me kɔ nyuie le kɔda si womenya o ŋu: woate ŋu atrɔ asi le ŋkuɖodzinu ƒe teƒe ɖesiaɖe si womate ŋu aɖo o ŋu, eye dɔwɔwɔ ɖesiaɖe si womenya o ate ŋu atrɔ nu ɖesiaɖe si woɖo ɖe heap-allocated. Nusiwo wotsɔ sɔ kple wo nɔewo wògbɔ eme sia kpɔa dedienɔnɔ ta le esi wògblẽa nu le nusiwo sɔ pɛpɛpɛ vie ta.

Nukatae scalability nye kuxi alea le practical points-to analysis me?

Gbeɖiɖi ƒe dzesiwo me dzodzro xɔ asi le dzɔdzɔme nu elabena ele be wòalé ŋku ɖe ƒomedodo siwo le ɖoɖowɔɖi blibo aɖe me ŋu le ɣeyiɣi ɖeka me. Le C codebase gã siwo mede blibo o gome la, point-to sets ate ŋu adzi ɖe edzi le agbɔsɔsɔ gã me abstraction manɔmee. Numekulawo kpɔa esia gbɔ kple mɔnu siwo wotu ɖe didi dzi, nɔnɔmetata siwo mebɔ o, kple nya siwo ƒo xlãe ƒe seselelãmemanɔamesi ƒe asitsatsa — lɔlɔ̃ ɖe nusiwo bu ƒe nusiwo sɔ pɛpɛpɛ dzi ɖe numekuku ɣeyiɣi siwo gakpɔtɔ nye nusi woate ŋu awɔ le xexeame ŋutɔŋutɔ ƒe sededewo dzi.

Aleke Mewayz kpena ɖe asitsalawo ŋu be woaƒo asa na dɔwɔwɔ ɖe nyatakaka siwo mede blibo o dzi?

Mewayz ƒoa asitsadɔ siwo wu 207 nu ƒu — tso CRM kple e-commerce dzi va ɖo numekuku, ƒuƒoƒo dzikpɔkpɔ, kple nyatakakawo ƒe ɖoɖowɔwɔ dzi — ɖe mɔnu ɖeka dzi. Esia ɖea nyatakakawo ƒe mama si zia asitsalawo dzi be woawɔ nyametsotso le nyatakaka siwo ƒe akpa aɖewo ŋu la ɖa. Esi ame 138,000+ zãlawo kple ɖoɖowo dze egɔme tso $19/ɣleti dzi ta la, Mewayz naa dɔwɔwɔ ƒe blibodede si asitsatsa ŋuti nyametsotso nyuiwo bia.

Èle klalo be yeaɖe blibomademade ɖa le wò asitsadɔwo mea? Dze wò Mewayz akɔnta gɔme egbea le app.mewayz.com eye nàkpɔ nusi wòfia be nàwɔ wò dɔwɔƒe bliboa tso mɔ̃ ɖeka si me gbeɖiɖi le, si wɔ ɖeka dzi.