Lectures will be recorded for remote students through Zoom and can be accessed through Panopto. Old lecture videos from 2012 are also available: 2012 CS1760 Lecture Videos. Date Topic Slides Recording
ä»å㯠"Double-Checked Locking" (以ä¸DCL)ã¨ãããã«ãã¹ã¬ããããã°ã©ã åãã®ã¤ãã£ãªã ãä¾ã«ãã¦ãC++0xã®(ä½ã¬ã¤ã¤åã)ãã«ãã¹ã¬ããæ©è½ã®å©ç¨æ¹æ³ãç´¹ä»ãã¦ã¿ã¾ãã DCLã¨ã¯ããããã¯âæ¡ä»¶å¤å®ãã¨ãããã¸ãã¯ããæ¡ä»¶å¤å®âããã¯â(å度)æ¡ä»¶å¤å®ãã¨æ¸ãæããã¤ãã£ãªã ã§ã主ã«é 延åæåãªã©ã®å¦çã«ããã¦ããã¯ã®ãªã¼ãã¼ããããæ¸ããããã«ç¨ãããã¾ããDCLã¯ã·ã³ãã«ãã¤å¹æã®é«ãã¤ãã£ãªã ã ã£ãã®ã§ãä¸ææãã¦ã¯ãããã¾ãããã¨ããããDCLã¯ã³ã³ãã¤ã©ãCPUã«ãããªãªã¼ãã¼ã®å½±é¿ã«ããæ£ããåä½ããªãå ´åããããã¨ãããã£ããã(åè1ãåè2)ãä»ã§ã¯ã¢ã³ããã¿ã¼ã³ã¨å¼ã°ãããã¨ããããå§æ«ã§ãã ããããDCLã®åé¡ç¹ã¯ãã¡ã¢ãªã¢ãã«ã«é¢ããç¥èããã¾ãç¥ããã¦ããªãã£ããã¨ã¨ãããã°ã©ãã³ã°è¨èªã®ä»æ§ã§ã¡ã¢ãªã¢ãã«ãæ£ããå®ç¾©ããã¦
Spinlocks and Read-Write Locks Most parallel programming in some way will involve the use of locking at the lowest levels. Locks are primitives that provide mutual exclusion that allow data structures to remain in consistent states. Without locking, multiple threads of execution may simultaneously modify a data structure. Without a carefully thought out (and usually complex) lock-free algorithm, t
A benchmark for a variety of spin-lock implementation using x86 primitives. These variations are as follows: * lock_none.h To provide baseline perfor mance compared to other implementation. * lock_tas.h Naive spin-lock implementaiont with test-and-set('xchg'). * lock_tas_nopause.h Based on lock_tas.h, 'pause' is removed. * lock_tas_mov.h Based on lock_tas.h, lock_release() is improved with 'mov' t
ãªãªã¼ã¹ãé害æ å ±ãªã©ã®ãµã¼ãã¹ã®ãç¥ãã
ææ°ã®äººæ°ã¨ã³ããªã¼ã®é ä¿¡
å¦çãå®è¡ä¸ã§ã
j次ã®ããã¯ãã¼ã¯
kåã®ããã¯ãã¼ã¯
lãã¨ã§èªã
eã³ã¡ã³ãä¸è¦§ãéã
oãã¼ã¸ãéã
{{#tags}}- {{label}}
{{/tags}}