Para-virtualized high-performance computing and GDI acceleration

A computer system, comprising:

• one or more processors;
• and
• one or more hardware storage devices having stored thereon computer-executable instructions that, when executed by the one or more processors, cause the computer system to provide GPU-accelerated computing functionality to a guest application executing in a child partition of a para-virtualized execution environment, including the following:
• an act of instantiating a virtual machine session, including instantiating a hypervisor that provides (i) a root partition having access to the physical GPU, and (ii) the child partition which executes the guest operating system;
• an act of presenting a graphics device interface (“GDI”) interface to the guest operating system, the GDI interface executing within a user-mode of the child partition, the GDI interface exposing one or more GDI command interfaces to a graphics runtime associated with the guest operating and forwarding GDI accelerated commands from the graphics runtime to a kernel-mode driver (“KMD”) of a virtualized graphics processing unit (“vGPU”) executing within the child partition;
• an act of presenting the vGPU to the guest operating system, including:
• presenting an application programming interface of a user-mode driver (“UMD”) that executes within the user-mode of the child partition and that presents a plurality of computer shader device driver interfaces (“DDIs”) to the guest operating system, the plurality of compute shader DDIs enabling the guest operating system to send compute shader commands to the vGPU for performing general-purpose graphics processing unit (“GPGPU”) computations at the physical GPU using a compute shader;
• and
• presenting an application programming interface of a kernel-mode driver (“KMD”) that executes in a kernel mode of the child partition and that includes:
• a GDI component that implements one or more GDI command interfaces for receiving one or more accelerated GDI operations from the GDI interface for execution on the physical GPU;
• and
• a compute shader component that receives a command buffer from the UMD and that constructs a corresponding direct memory access (“DMA”) buffer;
• an act of a render component executing within the root partition receiving a physical GPU-specific compute shader command and a GDI acceleration rendering command from the KMD vGPU;
• an act of the render component scheduling the physical GPU-specific compute shader command for execution at the physical GPU;
• and
• an act of the render component scheduling the GDI acceleration rendering command on a GDI composition device within the root partition, the GDI composition device being configured to execute the GDI acceleration rendering command at the physical GPU, the GDI composition device also being configured to mark a GDI surface corresponding to the GDI acceleration rendering command as sharable for composition by a desktop.

A computer program product for use at a computer system, the computer program product for implementing a method for providing GPU-accelerated computing functionality to a guest operating system executing in a child partition of a para-virtualized execution environment, the computer program product comprising one or more hardware storage devices having stored thereon computer-executable instructions that, when executed at a processor, cause the computer system to perform the method, including the following:

• instantiate a virtual machine session, including instantiating a hypervisor that provides (i) a root partition having access to a physical GPU, and (ii) the child partition which executes the guest operating system;
• present a graphics device interface (“GDI”) interface to the guest operating system, the GDI interface executing within a user-mode of the child partition, the GDI interface exposing one or more GDI command interfaces to a graphics runtime associated with the guest operating and forwarding GDI accelerated commands from the graphics runtime to a kernel-mode driver (“KMD”) of a virtualized graphics processing unit (“vGPU”) executing within the child partition;
• present the vGPU to the guest operating system, including:
• presenting an application programming interface of a user-mode driver (“UMD”) that executes within the user-mode of the child partition and that presents a plurality of compute shader device driver interfaces (“DDIs”) to the guest operating system, the plurality of compute shader DDIs enabling the guest operating system to send compute shader commands to the vGPU for performing general-purpose graphics processing unit (“GPGPU”) computations at the physical GPU using a compute shader;
• and
• presenting an application programming interface of a kernel-mode driver (“KMD”) that executes in a kernel mode of the child partition and that includes:
• a GDI component that implements one or more GDI command interfaces for receiving one or more accelerated GDI operations from the GDI interface for execution on the physical GPU;
• and
• a compute shader component that receives a command buffer from the UMD and that constructs a corresponding direct memory access (“DMA”) buffer;
• receive, at a render component executing within the root partition, at least one physical GPU-specific compute shader command and at least one GDI acceleration rendering command from the KMD of the vGPU;
• schedule the at least one physical GPU-specific compute shader command for execution at the physical GPU;
• and
• schedule the at least one GDI acceleration rendering command on a GDI composition device within the root partition, the GDI composition device being configured to execute the GDI acceleration rendering command at the physical GPU, the GDI composition device also being configured to mark a GDI surface corresponding to the GDI acceleration rendering command as sharable for composition by a desktop.