{"id":116280,"date":"2024-06-02T15:40:23","date_gmt":"2024-06-02T13:40:23","guid":{"rendered":"https:\/\/prohoster.info\/blog\/novosti-interneta\/novaya-optimizacziya-funkczii-memset-v-glibc"},"modified":"2024-06-02T15:40:23","modified_gmt":"2024-06-02T13:40:23","slug":"novaya-optimizacziya-funkczii-memset-v-glibc","status":"publish","type":"post","link":"https:\/\/prohoster.info\/ro\/blog\/news\/novaya-optimizacziya-funkczii-memset-v-glibc","title":{"rendered":"Nou\u0103 optimizare a func\u021biei memset() \u00een glibc","gt_translate_keys":[{"key":"rendered","format":"text"}]},"content":{"rendered":"
Inginerul de la Intel, Noah Goldstein, a optimizat func\u021bia memset() din biblioteca glibc. Aceast\u0103 optimizare ofer\u0103 un c\u00e2\u0219tig de performan\u021b\u0103 de aproximativ 7,5% pe versiunile desktop ale procesoarelor din arhitecturile Skylake-X \u0219i Ice Lake. Pentru versiunile server, c\u00e2\u0219tigul de performan\u021b\u0103 este u\u0219or mai mic, \u00een principal din cauza performan\u021bei generale mai reduse a unui singur nucleu.<\/p>\n
\u00cen implementarea anterioar\u0103 a func\u021biei memset() a fost folosit\u0103 instruc\u021biunea de asamblare rep stosb. P\u00e2n\u0103 de cur\u00e2nd, aceast\u0103 instruc\u021biune a func\u021bionat suficient de repede, datorit\u0103 optimiz\u0103rii interne zero-over-zero writeback. Totu\u0219i, \u00een aceast\u0103 optimizare a fost descoperit\u0103 o poten\u021bial\u0103 vulnerabilitate, care poate conduce la un atac prin canale laterale. Ca urmare, optimizarea zero-over-zero writeback a fost anulat\u0103, ceea ce a dus la o deteriorare a performan\u021bei rep stosb. \u00cen noua versiune a memset(), instruc\u021biunea rep stosb este folosit\u0103 \u00een continuare, dar sub condi\u021bii mai stricte.<\/p>\n
Ceea ce s-a schimbat poate fi \u00een\u021beles prin modificarea urm\u0103torului comentariu din cod, care descrie detaliile implement\u0103rii memset()<\/p>\n
Versiunea anterioar\u0103 a descrierii:<\/p>\n
\/* memset is implemented as: 1. Use overlapping store to avoid branch. 2. If size is less than VEC, use integer register stores. 3. If size is from VEC_SIZE to 2 * VEC_SIZE, use 2 VEC stores. 4. If size is from 2 * VEC_SIZE to 4 * VEC_SIZE, use 4 VEC stores. 5. On machines ERMS feature, if size is greater or equal than __x86_rep_stosb_threshold then REP STOSB will be used. 6. If size is more to 4 * VEC_SIZE, align to 4 * VEC_SIZE with 4 VEC stores and store 4 * VEC at a time until done. *\/ <\/p>\n
Noua versiune a descrierii:<\/p>\n
\/* memset is implemented as: 1. Use overlapping store to avoid branch. 2. If size is less than VEC, use integer register stores. 3. If size is from VEC_SIZE to 2 * VEC_SIZE, use 2 VEC stores. 4. If size is from 2 * VEC_SIZE to 4 * VEC_SIZE, use 4 VEC stores. 5. If size is more to 4 * VEC_SIZE, align to 1 * VEC_SIZE with 4 VEC stores and store 4 * VEC at a time until done. 6. On machines ERMS feature, if size is range \t [__x86_rep_stosb_threshold, __x86_memset_non_temporal_threshold) \t then REP STOSB will be used. 7. If size >= __x86_memset_non_temporal_threshold, use a \t non-temporal stores. *\/<\/p>\n