摘要:
A method and apparatus are disclosed for providing a DMA process. Accordingly, a DMA process is initiated for moving data from contiguous first locations to contiguous second locations and to a third location or third locations. Within the DMA process the data from each of the contiguous first locations is retrieved and stored in a corresponding one of the contiguous second locations and in the third location or corresponding one of the third locations. The DMA process is performed absent retrieving the same data a second other time prior to storing of same within the corresponding one of the contiguous second locations and in the third location or corresponding one of the third locations.
摘要:
A method and apparatus are disclosed for prefragmenting or presegmenting data. A data protocol type and length are provided. Based on the type and length, a linked list memory buffer structure is determined. Locations within the memory buffer structure for storage of data therein is also determined allowing for sufficient unused memory therein to support one of a fragmentation operation and a segmentation operation subsequent to memory allocation. As such, further memory allocation during a fragmentation operation or a segmentation operation is obviated. The determined memory buffer structure is then allocated.
摘要:
A method and apparatus are disclosed for providing a DMA process. Accordingly, a DMA process is initiated for moving data from contiguous first locations to contiguous second locations and to a third location or third locations. Within the DMA process the data from each of the contiguous first locations is retrieved and stored in a corresponding one of the contiguous second locations and in the third location or corresponding one of the third locations. The DMA process is performed absent retrieving the same data a second other time prior to storing of same within the corresponding one of the contiguous second locations and in the third location or corresponding one of the third locations.
摘要:
A method and apparatus are disclosed for providing a DMA process. Accordingly, a DMA process is initiated for moving data from contiguous first locations to contiguous second locations and to a third location or third locations. Within the DMA process the data from each of the contiguous first locations is retrieved and stored in a corresponding one of the contiguous second locations and in the third location or corresponding one of the third locations. The DMA process is performed absent retrieving the same data a second other time prior to storing of same within the corresponding one of the contiguous second locations and in the third location or corresponding one of the third locations.
摘要:
A method and apparatus are disclosed for prefragmenting or presegmenting data. A data protocol type and length are provided. Based on the type and length, a linked list memory buffer structure is determined. Locations within the memory buffer structure for storage of data therein is also determined allowing for sufficient unused memory therein to support one of a fragmentation operation and a segmentation operation subsequent to memory allocation. As such, further memory allocation during a fragmentation operation or a segmentation operation is obviated. The determined memory buffer structure is then allocated.
摘要:
An architecture for use in packet processing and supporting compatibility with current BSD implementations for packet processing is proposed wherein two MBUF formats are supported. A first format includes a header portion and a data portion for storing data therein. A second format includes a header portion but is absent a data portion and is for addressing data stored within a cluster and external to the MBUF itself.
摘要:
A method and apparatus are disclosed for providing a DMA process. Accordingly, a DMA process is initiated for moving data from contiguous first locations to contiguous second locations and to a third location or third locations. Within the DMA process the data from each of the contiguous first locations is retrieved and stored in a corresponding one of the contiguous second locations and in the third location or corresponding one of the third locations. The DMA process is performed absent retrieving the same data a second other time prior to storing of same within the corresponding one of the contiguous second locations and in the third location or corresponding one of the third locations.
摘要:
An architecture for use in packet processing and supporting compatibility with current BSD implementations for packet processing is proposed wherein two MBUF formats are supported. A first format includes a header portion and a data portion for storing data therein. A second format includes a header portion but is absent a data portion and is for addressing data stored within a cluster and external to the MBUF itself.