Lines Matching refs:operations

49 Read-Write (RW) access operations by several lcores to the same memory area can generate a lot of d…
77 meaning that if all memory access operations are done on the first channel only, there is a potenti…
105 Performance is greatly improved when using bulk access operations.
171 operations in the DPDK environment.
176 On x86, atomic operations imply a lock prefix before the instruction,
188 DPDK generic rte_atomic operations are implemented by __sync builtins. These
193 So replacing the rte_atomic operations with __atomic builtins might improve
201 Some use cases require atomicity alone, the ordering of the memory operations
212 For example, the memory operations before the spinlock lock are allowed to
213 move to the critical section, but the memory operations in the critical section
216 On the other hand, the memory operations after the spinlock unlock are allowed
217 to move to the critical section, but the memory operations in the critical