TAILIEUCHUNG - Parallel Programming: for Multicore and Cluster Systems- P37

Parallel Programming: for Multicore and Cluster Systems- P37: Innovations in hardware architecture, like hyper-threading or multicore processors, mean that parallel computing resources are available for inexpensive desktop computers. In only a few years, many standard software products will be based on concepts of parallel programming implemented on such hardware, and the range of applications will be much broader than that of scientific computing, up to now the main application area for parallel computing | 352 6 Thread Programming Besides the explicit flush construct there is an implicit flush at several points of the program code which are a barrier construct entry to and exit from a critical region at the end of a parallel region at the end of a for sections or single construct without nowait clause entry and exit of lock routines which will be introduced below . Locking Mechanism The OpenMP runtime system also provides runtime library functions for a synchronization of threads with the locking mechanism. The locking mechanism has been described in Sect. and in this chapter for Pthreads and Java threads. The specific locking mechanism of the OpenMP library provides two kinds of lock variables on which the locking runtime routines operate. Simple locks of type om loctt can be locked only once. Nestable locks of type ompnestdoctt can be locked multiple times by the same thread. OpenMP lock variables should be accessed only by OpenMP locking routines. A lock variable is initialized by one of the following initialization routines void omp_init_lock om loctt lock void omp_init_nest_lock om _nesLlockt lock for simple and nestable locks respectively. A lock variable is removed with the routines void ompdestroylock om loctt lock void omp_destroy_nest_lock om _nesLloctt lock . An initialized lock variable can be in the states locked or unlocked. At the beginning the lock variable is in the state unlocked. A lock variable can be used for the synchronization of threads by locking and unlocking. To lock a lock variable the functions void ompseLlock om lockt lock void ompseLnestlock omp_nest_loc ct lock are provided. If the lock variable is available the thread calling the lock routine locks the variable. Otherwise the calling thread blocks. A simple lock is available when no other thread has locked the variable before without unlocking it. A nestable lock variable is available when no other thread has locked the variable without unlocking it or when the calling .

• Phần cứng
• sửa chữa máy tính
• sổ tay phần cứng
• MainBoard
• thiết bị máy tính
• thiết bị mạng
• lắp ráp máy tính
• Phần cứng máy tính
• lỗi máy tính
• hệ thống mạch máy tính
• điện máy tính
• Tài liệu máy tính