{"id":121189,"date":"2024-12-31T15:46:00","date_gmt":"2024-12-31T13:46:00","guid":{"rendered":"https:\/\/prohoster.info\/blog\/novosti-interneta\/kollizii-v-sokrashhyonnyh-identifikatorah-kommitov-v-yadre-linux"},"modified":"2024-12-31T15:46:00","modified_gmt":"2024-12-31T13:46:00","slug":"kollizii-v-sokrashhyonnyh-identifikatorah-kommitov-v-yadre-linux","status":"publish","type":"post","link":"https:\/\/prohoster.info\/ro\/blog\/news\/kollizii-v-sokrashhyonnyh-identifikatorah-kommitov-v-yadre-linux","title":{"rendered":"Coliziunile \u00een identificatorii scurta\u021bi ai comitelor \u00een nucleul Linux","gt_translate_keys":[{"key":"rendered","format":"text"}]},"content":{"rendered":"
Kees Cook, fost administrator principal al sistemului la kernel.org \u0219i lider al echipei de securitate Ubuntu, a demonstrat posibilitatea cre\u0103rii unui commit al c\u0103rui ID scurt coincide cu un commit anterior ad\u0103ugat \u00een nucleul Linux. Experimentul a fost realizat ca o dovad\u0103 a fezabilit\u0103\u021bii tranzi\u021biei la ID-uri scurte de commit de 16 caractere \u00een nucleul Linux, despre care s-a discutat anterior pe lista de discutii a dezvoltatorilor nucleului, dar care nu a fost aprobat de Linus Torvalds. <\/p>\n
ID-urile scurte ale commit-urilor sunt generate prin p\u0103strarea primelor 12 caractere din hash-ul SHA-1 (48 de bi\u021bi din 160 de bi\u021bi). Av\u00e2nd \u00een vedere c\u0103 num\u0103rul de obiecte \u00een nucleu identificate prin hash-ul SHA-1 a dep\u0103\u0219it 13 milioane, apari\u021bia coliziunilor utiliz\u00e2nd un prefix de 12 caractere a devenit o chestiune de timp. Ca exemplu, sunt prezentate deja obiecte ad\u0103ugate \u00een nucleu care se suprapun prin ID-urile lor de 11 caractere. \u00cen plus, s-a men\u021bionat c\u0103 suprapunerea ID-urilor de 12 caractere a fost deja \u00eenregistrat\u0103 \u00een luna octombrie, dar \u00eenainte de trimiterea patch-ului, utilitarul checkpatch a descoperit problema. <\/p>\n
Identificatorii scurta\u021bi sunt utiliza\u021bi la publicarea linkurilor scurte pentru commit-uri, precum \u0219i atunci c\u00e2nd se trimite modific\u0103ri \u00een eticheta \u00abFixes\u00bb, ca un link c\u0103tre commit-ul care a rezolvat problema \u00een patch-ul trimis (de exemplu, \u00abFixes: e21d2170f366\u00bb). Apari\u021bia coliziunilor, \u00een care mai multe modific\u0103ri diferite sunt legate de un singur identificator scurt, poate duce la disfunc\u021bionalit\u0103\u021bi \u00een instrumentele de analiz\u0103 \u0219i verificare a modific\u0103rilor care iau \u00een considerare con\u021binutul etichetei \u00abFixes\u00bb. De exemplu, aceste etichete sunt luate \u00een considerare \u00een handler-ul check_fixes, utilizat \u00een ramura linux-next, precum \u0219i \u00een scripturile de analiz\u0103 a corec\u021biilor de vulnerabilitate \u0219i monitorizarea ciclului de via\u021b\u0103 al patch-urilor. <\/p>\n
Linus Torvalds has expressed skepticism about the proposal to increase the minimum size of shortened identifiers, as the actual number of commits in the repository is significantly lower than the number of objects (approximately 1\/8). Most likely, if random collisions occur, they will be between a commit and an object of a different type (for example, a blob or a branch). In his opinion, shortened identifiers are meant to be clear, readable, and easily quotable, and there are currently no objective reasons to increase their size. <\/p>\n
One of the developers suggested achieving a reduction in size while increasing the number of significant bits by using a new format based on Base36 encoding (characters 0-9a-z) instead of hexadecimal digits. According to Linus, such a change would create more problems than it solves. For example, support for the new format would need to be added to existing utilities, and a format identifier would need to be introduced to distinguish between the old and new formats. <\/p>\n
To demonstrate that the issue with shortened identifiers is not theoretical and its resolution should not be delayed, Kees Cook prepared a documentation change for the kernel, the shortened identifier of which (1da177e4c3f4) matched the commit identifier for creating the kernel branch 2.6.12-rc2. The collision was found after 6 hours of calculations on a system with an NVIDIA GeForce RTX 3080 GPU. <\/p>\n
Selectarea a fost efectuat\u0103 cu utilizarea instrumentului lucky-commit \u2014 \u00een textul patch-ului \u021bint\u0103 au fost ad\u0103ugate spa\u021bii aleatorii p\u00e2n\u0103 c\u00e2nd prefixul SHA-1 de 12 caractere a coincis cu prefixele commit-urilor deja existente \u00een nucleu. Potrivit lui Case, problema nu o constituie at\u00e2t intersec\u021biile aleatorii, c\u00e2t posibilitatea manipul\u0103rii identificatorilor scurta\u021bi \u00een scopuri mali\u021bioase, de exemplu, pentru a ocolii anumite verific\u0103ri. \n Sursa: opennet.ro<\/a> <\/p>","protected":false,"gt_translate_keys":[{"key":"rendered","format":"html"}]},"excerpt":{"rendered":"