公开(公告)号：US20060227795A1

公开(公告)日：2006-10-12

申请号：US11156070

申请日：2005-06-17

Applicant: Amit Bande ,  Krishna Mohan I.V. ,  Sreekumar Seshadri

Inventor： Amit Bande ,  Krishna Mohan I.V. ,  Sreekumar Seshadri

IPC: H04J3/16 ,  H04J3/22 ,  H04L12/56 ,  H04L12/28

CPC classification number: H04L49/9026 ,  H04L49/90

Abstract: Techniques are described for reading a plurality of data packets, which can be communicated over a network, without sacrificing the efficiency of network operation, and for reading data packets communicated over a network without unnecessary memory copies. A receive buffer of a first size, which is at least twice a maximum size of data packets in the plurality of data packets, is used. A first amount of data is read into the receive buffer and the size of a last packet contained within the first amount of data is determined. The size of an unread portion of the last read packet is determined and a second amount of data, which is equal to size of an unread portion of the last read packet, is then read into the receive buffer.

Abstract translation: 描述了用于读取可以通过网络传送的多个数据分组的技术，而不牺牲网络操作的效率，以及用于读取通过网络传送的数据分组，而不需要不必要的存储器拷贝。 使用第一大小的接收缓冲器，其为多个数据分组中的数据分组的最大大小的至少两倍。 将第一数据量读入接收缓冲器，并且确定包含在第一数据量内的最后一个分组的大小。 确定最后一个读取分组的未读部分的大小，然后将等于最后读取分组的未读部分的大小的第二数据量读入接收缓冲器。