Host context techniques for server based graphics processing

• US 9,613,390 B2
• Filed: 10/29/2013
• Issued: 04/04/2017
• Est. Priority Date: 05/02/2012
• Status: Active Grant

• Current Assignee: NVIDIA Corporation
• Original Assignee: NVIDIA Corporation
• Inventors: Diard, Franck
• Primary Examiner(s): McDowell, Jr., Maurice L
• Assistant Examiner(s): Ricks, Donna J
• Application Number:
  US14/066,673
• Publication Number:
  US 20140055470A1
• Time in Patent Office: 1,253 Days
• Field of Search: None
• CPC Class Codes: G06F 13/10 Program control for peripheral devices G06F13/14 - G06F13/42 take precedence
• CPC Class Codes: G06F 9/44 Arrangements for executing specific programs
• CPC Class Codes: G06F 9/455 Emulation; Interpretation; Software simulation, e.g. virtualisation or emulation of application or operating system execution engines
• CPC Class Codes: G06F 9/45545 Guest-host, i.e. hypervisor is an application program itself, e.g. VirtualBox
• CPC Class Codes: G06F 9/545 where tasks reside in different layers, e.g. user- and kernel-space
• CPC Class Codes: G06T 1/20 Processor architectures; Processor configuration, e.g. pipelining

Abstract

The server based graphics processing techniques, describer herein, include receiving function calls by a three dimension graphics application programming interface host-guest communication manager (D3D HGCM) service module from one or more given instances of a guest shin layer through a communication channel of a host-guest communication manager (HGCM). The one or more given instances of the guest shim layer are executing under control of a respective given instance of a guest operating system. The HGCM and D3D HGCM service module are executing under control of a host operating system. The rendering context for each function call received from the each instance of the guest shim layer is determined by the D3D HGCM service module. Each function call of a given rendering context is sent by the D3D HGCM service module to a corresponding device specific kernel mode driver of a given graphics processing unit for scheduling execution with the given graphics processing unit of the given rendering context.

Claims

1. A method comprising:

delivering hardware resources, software and information of a computing platform to a plurality of user devices as a utility delivered over a network, wherein rendering display frames output on a display of a given user device by the hardware resources, software and information on the computing platform includes

receiving function calls by a three dimension graphics application programming interface host-guest communication manager (D3D HGCM) service module from one or more given instance of a guest shim layer through a communication channel of a host-guest communication manager (HGCM), wherein the one or more given instances of the guest shim layer are executing under control of a respective given instance of a guest operating system and wherein the HGCM and D3D HGCM service module me executing under control of a host operating system;

determining, by the D3D HGCM service module, a rendering context for each function call received from the each instance of the guest shim layer;

sending, by the D3D HGCM service module, each function call of a given rendering context to a corresponding thunk layer;

sending each function call of the given rendering context from the corresponding thunk layer to a corresponding operating system kernel mode driver;

sending each function call of the given rendering context from the operating system kernel mode driver to a corresponding device specific kernel mode driver of a given graphics processing unit for the given rendering context;

scheduling, by the corresponding device specific kernel mode driver, each function call of the given rendering context including setting command registers of the given graphics processing unit to execute each function call of the given rendering context.

2. The method according, to claim 1, wherein the rendering context is determined from a toke identifier.

3. The method according to claim 1, wherein the rendering context is determined from a handle to a data structure that defines the rendering context of a corresponding guest.

4. The method according to claim 1, further comprising routing rendered frame data to an encoder by the D3D HGCM service module for a display related function call.

5. The method according to claim 4, further comprising:

compressing, the rendered frame data by the encoder; and

sending the compressed rendered frame data from the encoder through a network interface to a user device.

6. The method according to claim 1, wherein the guest shim layer is not a display device interface user mode driver of a virtual machine manager including the HGCM.

7. One or more non-transitory computing device readable media storing one or more sets of computing device executable instructions which when executed by one or more processing units perform a process comprising:

receiving function calls by a three dimension graphics application programming interface host-guest communication manager (D3D HGCM) service module from one or more given instances of a guest shim layer through a communication channel of a host-guest communication manager (HGCM), wherein the one or more given instances of the guest shim layer are executing under control of a respective given instance of a guest operating system and wherein the HGCM and D3D HGCM service module are executing under control of a host operating system;

determining, by the D3D HGCM service module, a rendering context for each function call received from the each instance of the guest shim layer;

sending, by the D3D HGCM service module, each function call of a given rendering context to a corresponding thunk layer;

sending each function call of the given rendering context from the corresponding thunk layer to a corresponding operating system kernel mode driver;

sending each function call of the given rendering context from the operating system kernel mode driver to a corresponding device specific kernel mode driver of a given graphics processing unit for the given rendering context;

scheduling, by the corresponding device specific kernel mode driver, each function call of the given rendering context including setting command registers of the given graphics processing unit to execute each function call of the given rendering context.

8. The one or more non-transitory computing device readable media according to claim 7, wherein the guest shim layer is not a display device interface user mode driver of a virtual machine manager including the HGCM.

9. The one or more non-transitory computing device readable media according to claim 8, wherein the rendering, context is determined from a take identifier.

10. The one or more non-transitory computing device readable media according to claim 8, wherein the rendering context is determined from a handle to a data structure that defines the rendering context of a corresponding guest.

11. The one or more non-transitory computing device readable media according to claim 8, further comprising routing rendered frame data to an encoder by the D3D HGCM service module for a display related function call.

12. The one or more non-transitory computing device readable media according to claim 11, further comprising:

compressing, the rendered frame data by the encoder; and

sending the compressed rendered frame data from the encoder through a network interface to a user device.

13. The one or more non-transitory computing device readable media according to claim 7, wherein the HGCM comprises a software development kit (SDK) of VirtualBox.

14. The one or more non-transitory computing device readable media according to claim 7, wherein the guest shim layer comprises a user mode driver dynamic linked library.