Container access to graphics processing unit resources
Systems and methods for container access to graphics processing unit (GPU) resources are disclosed herein. In some embodiments, a computing system may include a physical GPU and kernel-mode driver circuitry, to communicatively couple with the physical GPU to create a plurality of emulated GPUs and a corresponding plurality of device nodes. Each device node may be associated with a single corresponding user-side container to enable communication between the user-side container and the corresponding emulated GPU. Other embodiments may be disclosed and/or claimed.
1. A computing system, comprising:
a physical graphics processing unit, GPU; and
kernel-mode driver circuitry, to communicatively couple with the physical GPU, to create a plurality of emulated GPUs and a corresponding plurality of device nodes, wherein each device node is to be associated with a single corresponding user-side container to enable communication between the user-side container and the corresponding emulated GPU;
wherein an emulated GPU is to support a first set of GPU features, the physical GPU supports a second set of GPU features, and the first set is a proper subset of the second set.
2. The computing system of claim 1 , wherein each of the emulated GPUs is to emulate the physical GPU.
3. The computing system of claim 1 , wherein each user-side container includes a user-mode driver to communicate with the corresponding emulated GPU via the corresponding device node.
4. The computing system of claim 3 , wherein each user-side container includes one or more software applications in communication with the user-mode driver, and each of the one or more software applications is to utilize a resource of the physical GPU.
5. The computing system of claim 1 , wherein the kernel-mode driver circuitry further includes a scheduler, and the scheduler is to communicate with the plurality of emulated GPUs and the physical GPU.
6. One or more non-transitory computer readable media having instruction thereon that, in response to execution by one or more processing devices of a computing system, cause the computing system to:
provide a kernel-mode driver to create a plurality of emulated graphics processing units, GPUs and a corresponding plurality of device nodes, wherein the plurality of emulated GPUs are communicatively coupled to a physical GPU; and
associate each device node with a single corresponding user-side container to enable communication between the user-side container and the corresponding emulated GPU to utilize resources of the physical GPU;
wherein a first emulated GPU does not emulate the physical GPU, the first emulated GPU is associated with a first device node, the kernel-mode driver is a first kernel-mode driver, and the instructions are further to, upon execution by the one or more processing devices of the computing system, cause the computing system to:
provide a second kernel-mode driver, arranged such that the second kernel-mode driver is disposed between a user-mode driver in a user-side container and the first kernel-mode driver in a communication pathway.
7. The one or more non-transitory computer readable media of claim 6 , wherein each of the emulated GPUs is to emulate the physical GPU.
8. The one or more computer readable media of claim 6 , wherein each user-side container includes a user-mode driver to communicate with the corresponding emulated GPU via the corresponding device node.
9. The one or more non-transitory computer readable media of claim 8 , wherein each user-side container includes one or more software applications in communication with the user-mode driver, and each of the one or more software applications is to utilize a resource of the physical GPU.
10. The one or more non-transitory computer readable media of claim 6 , wherein the instructions are further to, upon execution by the one or more processing devices of the computing system, cause the computing system to:
provide a scheduler to communicate with the plurality of emulated GPUs and the physical GPU.
11. The one or more non-transitory computer readable media of claim 10 , wherein the scheduler is to enforce a quality of service policy on use of the physical GPU by the emulated GPUs.
12. The one or more non-transitory computer readable media of claim 10 , wherein the scheduler is to enforce a secure isolation policy for the emulated GPUs.
13. The one or more non-transitory computer readable media of claim 6 , wherein the instructions are further to, upon execution by the one or more processing devices of the computing system, cause the computing system to:
partition a graphics memory among the plurality of emulated GPUs.
14. The one or more non-transitory computer readable media of claim 13 , wherein the graphics memory is to be utilized for pixel loading or as a write command buffer.
15. The one or more non-transitory computer readable media of claim 6 , wherein the instructions are further to, upon execution by the one or more processing devices of the computing system, cause the computing system to:
create a second emulated GPU communicatively coupled to the physical GPU; and
associate the second emulated GPU with a virtual machine to enable communication between the virtual machine and the corresponding emulated GPU.
16. The one or more non-transitory computer readable media of claim 15 , wherein the virtual machine includes a kernel-mode GPU driver and a user-mode GPU driver.
17. The one or more non-transitory computer readable media of claim 6 , wherein a first user-side container includes a user-mode driver to communicate with the second kernel-mode driver.
18. One or more non-transitory computer readable media having instruction thereon that, in response to execution by one or more processing devices of a computing system, cause the computing system to:
associate a user-side container with a kernel-side device node, wherein the kernel-side device node corresponds to an emulated graphics processing unit, GPU, of a kernel-mode driver and the emulated GPU is communicatively coupled to a physical GPU; and
allow one or more software applications in the user-side container to utilize a resource of the physical GPU via the kernel-side device node and the emulated GPU;
wherein the kernel-mode driver includes a display manager to couple with a frame buffer of the user-side container to selectively bring the frame buffer to a control path for the physical GPU.
19. The one or more non-transitory computer readable media of claim 18 , wherein each of the emulated GPUs is to emulate the physical GPU.
20. The one or more non-transitory computer readable media of claim 18 , wherein the kernel-mode driver includes a scheduler, and the scheduler is to communicate with the emulated GPU and the physical GPU.
21. The one or more non-transitory computer readable media of claim 20 , wherein the emulated GPU is to support a first set of GPU features, the physical GPU supports a second set of GPU features, and the first set is a proper subset of the second set.
22. The one or more non-transitory computer readable media of claim 18 , wherein:
the emulated GPU does not emulate the physical GPU;
the kernel-mode driver is a first kernel-mode driver;
as part of allowance of one or more software applications in the user-side container to utilize a resource of the physical GPU via the kernel-side device node and the emulated GPU, allow one or more software applications in the user-side container to utilize the resource of the physical GPU via the kernel-side device node, the emulated GPU, and a second kernel-mode driver; and
the device node is disposed between the second kernel-mode driver and the emulated GPU in a communication pathway.