ACCELERATOR BUFFER ACCESS

摘要

Technologies are generally described for methods and systems effective to provide accelerator buffer access. An operating system may allocate a range of addresses in virtual address spaces and a range of addresses in a buffer mapped region of a physical (or main) memory. A request to read from, or write to, data by an application may be read from, or written to, the virtual address space. A memory management unit may then map the read or write requests from the virtual address space to the main or physical memory. Multiple applications may be able to operate as if each application has exclusive access to the accelerator and its buffer. Multiple accesses to the buffer by application tasks may avoid a conflict because the memory controller may be configured to fetch data based on respective application identifiers assigned to the applications. Each application may be assigned a different application identifier.

主权项

1. A method for an accelerator to store data in a buffer, the method comprising:
receiving, a first request, by the accelerator, to process data associated with a first task dispatched by a first application; requesting a first data for the first task in a first range of addresses from a memory, wherein the first range of addresses are mapped to a second range of addresses in a first virtual space assigned to the first application; receiving the first data; storing the first data in the buffer; receiving, a second request, by the accelerator to process data associated with a second task dispatched by a second application; requesting for the second application at least some second data in a third range of addresses from the memory, wherein the third range of addresses is mapped to a fourth range of addresses in a second virtual space assigned to the second application; receiving a portion of the second data; and storing the portion of the second data in the buffer.