System and method for auditing buffer usage in a data processing system
    1.
    发明授权
    System and method for auditing buffer usage in a data processing system 失效
    在数据处理系统中审计缓冲区使用的系统和方法

    公开(公告)号:US6044477A

    公开(公告)日:2000-03-28

    申请号:US930507

    申请日:1998-01-26

    IPC分类号: G06F11/00 G06F11/34 G06F11/36

    摘要: A system and method are provided which allows the monitoring of the allocation and use of buffers allocated from a buffer pool to processes executing within a data processing system. Conventionally, if an instance of a process crashes or loses a reference to a buffer, that buffer is rendered inaccessible by other processes. Eventually, the buffer pool will become exhausted. Errors in a communication system are difficult to attribute to a single cause and, in particular within a communication stack, often arise as a consequence of a series of events. Typically, prior art communication monitoring systems only keep track of the latest process to have had access to a buffer and do not allow an investigation into the past history of the possession or access to the buffers. Accordingly, the present invention provides a system and method for monitoring the history of the possession of or access to buffers within a communication system. An audit vector is used to store indications of all of the processes which have had access to a buffer. In the event of a malfunction or at the instigation of a user the contents of the audit vector are output for investigation.

    摘要翻译: PCT No.PCT / GB96 / 00205 Sec。 371日期1998年1月26日 102(e)1998年1月26日PCT PCT 1996年1月31日PCT公布。 WO97 / 28501 PCT公开号 日期1997年8月7日提供了一种系统和方法,其允许监视从缓冲池分配的缓冲器的分配和使用到在数据处理系统内执行的处理。 通常,如果进程的实例崩溃或丢失对缓冲区的引用,则该缓冲区将被其他进程无法访问。 最后,缓冲池会耗尽。 通信系统中的错误难以归因于单一原因,特别是在通信堆栈内,通常是由于一系列事件而产生的。 通常,现有技术的通信监视系统仅跟踪最近进程以访问缓冲器,并且不允许调查过去拥有或访问缓冲器的历史。 因此,本发明提供了一种用于监视在通信系统内拥有或访问缓冲器的历史的系统和方法。 审计向量用于存储已访问缓冲区的所有进程的指示。 在发生故障或用户的唆使的情况下,输出审核向量的内容进行调查。