DYNAMIC PINNING OF VIRTUAL PAGES SHARED BETWEEN DIFFERENT TYPE PROCESSORS OF A HETEROGENEOUS COMPUTING PLATFORM

摘要

A computer system may support one or more techniques to allow dynamic pinning of the memory pages accessed by a non-CPU device, such as a graphics processing unit (GPU). The non-CPU may support virtual to physical address mapping and may thus be aware of the memory pages, which may not be pinned but may be accessed by the non-CPU. The non-CPU may notify or send such information to a run-time component such as a device driver associated with the CPU. The device driver may, dynamically, perform pinning of such memory pages, which may be accessed by the non-CPU. The device driver may even unpin the memory pages, which may be no longer accessed by the non-CPU. Such an approach may allow the memory pages, which may be no longer accessed by the non-CPU to be available for allocation to the other CPUs and/or non-CPUs.

主权项

1. A method to be performed in a computing platform having a plurality of heterogeneous processors, including at least a central processing unit (CPU) and a non-CPU, and a shared memory, the method comprising:
determining when a virtual memory page that a non-CPU is looking to access is not pinned; generating a request for pinning the virtual memory page; sending the request to a run-time component associated with the CPU; determining whether a total number of memory pages already pinned exceeds a threshold value; identifying one or more virtual memory pages pinned but no longer accessed; and unpinning the identified one or more virtual memory pages when the threshold value is exceeded.