公开(公告)号：US07603488B1

公开(公告)日：2009-10-13

申请号：US10892538

申请日：2004-07-15

申请人： Martin Gravenstein ,  Nirmalendu B. Patra ,  Andrew Probst ,  Dave Ohmann ,  Clair A. Hardesty

发明人： Martin Gravenstein ,  Nirmalendu B. Patra ,  Andrew Probst ,  Dave Ohmann ,  Clair A. Hardesty

IPC分类号： G06F13/28 ,  G06F3/00

CPC分类号： G06F12/0223 ,  G06F12/023 ,  G06F2212/171 ,  H04L1/1835 ,  H04L1/1874 ,  Y02D10/13

摘要： Systems and methods for providing efficient memory allocation, reduced processor intervention and power consumption, and increased memory access bandwidth. One embodiment comprises a system including a plurality of memory units which are accessible in parallel, a dynamic memory unit configured to dynamically allocate and deallocate storage space in the memory units, and a plurality of direct memory access (DMA) engines configured to access the memory units in parallel through the memory management subsystem. The system may be implemented in the MAC engine of a device that communicates with other devices via a wireless communication link. This embodiment may store packets in FIFOs within the memory units as elements of linked list data structures that can be joined together without having to move the previously stored data. DMA engines access a context table to obtain DMA channel information that enables them to move data through appropriate DMA channels.

摘要翻译： 提供高效内存分配，降低处理器干预和功耗以及增加内存访问带宽的系统和方法。 一个实施例包括包括可并行访问的多个存储器单元的系统，被配置为动态地分配和释放存储器单元中的存储空间的动态存储器单元，以及被配置为访问存储器的多个直接存储器访问（DMA）引擎 单元并行通过内存管理子系统。 该系统可以在经由无线通信链路与其他设备通信的设备的MAC引擎中实现。 该实施例可以将存储器单元中的FIFO中的分组存储为可以连接在一起而不必移动先前存储的数据的链表列表数据结构的元素。 DMA引擎访问上下文表以获取DMA通道信息，使其能够通过适当的DMA通道移动数据。