Dynamic virtual device failure recovery
Exemplary techniques for recovering from a graphics processor reset are herein disclosed. In an exemplary embodiment, a 3D graphics service manager 404 can detect that a graphics processing unit reset and can restart a rendering process configured to render 3D graphics for a virtual machine and cause a graphics buffer to be established between the rendering process and the virtual machine. In addition to the foregoing, other aspects are described in the detailed description, claims, and figures.
1. A computer system configured to recover from a graphics processor reset, comprising:
a processor;
a 3D graphics processing unit; and
a memory in communication with at least the processor when the computer system is operational, the memory having stored thereon computer readable instructions that upon execution cause the processor to:
establish a communication channel between a virtualization platform and a virtual machine operating on the same computer system;
detect that a rendering process configured to render graphics for the virtual machine on the 3D graphics processing unit terminated;
restart the rendering process;
send, via the communication channel, a signal indicating that the rendering process is being restarted to a virtual device driver executing in the virtual machine; and
establish a graphics buffer between the virtual device driver executing in the virtual machine and the restarted rendering process.
2. The computer system of claim 1 , wherein the computer readable instructions further comprise computer readable instructions that upon execution cause the processor to:
resend graphics data to render a full frame via the graphics buffer.
3. The computer system of claim 1 , wherein the computer readable instructions that upon execution cause the processor to restart the rendering process further comprise computer readable instructions that upon execution cause the processor to:
restart the rendering process in response to determining that the virtual machine was not conducting a denial of service attack before the rendering process terminated.
4. The computer system of claim 1 , wherein the computer readable instructions that upon execution cause the processor to restart the rendering process further comprise computer readable instructions that upon execution cause the processor to:
restart the rendering process in response to determining that an average latency of direct memory access transactions initiated by the virtual device driver before the rendering process terminated is less than a threshold.
5. The computer system of claim 1 , wherein the computer readable instructions that upon execution cause the processor to restart the rendering process further comprise computer readable instructions that upon execution cause the processor to:
restart the rendering process in response to determining that an amount of time the 3D graphics processing unit takes to respond to commands before the rendering process terminated is less than a threshold.
6. The computer system of claim 1 , wherein the computer readable instructions that upon execution cause the processor to restart the rendering process further comprise computer readable instructions that upon execution cause the processor to:
restart the rendering process in response to determining that an estimated amount of available graphics memory controlled by the 3D graphics processing unit before the rendering process terminated is greater than a threshold.
7. The computer system of claim 1 , wherein the computer readable instructions that upon execution cause the processor to restart the rendering process further comprise computer readable instructions that upon execution cause the processor to:
restart the rendering process and assign the rendering process to render on a second 3D graphics processing unit.
8. A computer-readable storage device including computer readable instructions for recovering from a 3D graphics processing unit reset, the computer-readable storage medium comprising instructions that upon execution by a processor cause the processor to:
establish a communication channel between a host operating system and a virtual machine;
run a rendering process in the host operating system, wherein the rendering process is configured to render graphics for the virtual machine on a 3D graphics processing unit;
detect that the rendering process terminated;
restart the rendering process;
send, via the communication channel, a message indicating that the rendering process was restarted;
establish, by a virtual device driver in the virtual machine, a graphics buffer in a range of guest memory; and
send, via the communication channel, 3D graphics data to the restarted rendering process.
9. The computer-readable storage device of claim 8 , wherein the computer readable instructions that upon execution cause the processor to:
resend graphics data to render a full frame via the graphics buffer.
10. The computer readable storage device of claim 8 , wherein the computer readable instructions that upon execution cause the processor to restart the rendering process further comprise computer readable instructions that upon execution cause the processor to:
restart the rendering process in response to determining that the virtual machine was not conducting a denial of service attack before the rendering process terminated.
11. The computer-readable storage device of claim 8 , wherein the computer readable instructions that upon execution cause the processor to restart the rendering process further comprise computer readable instructions that upon execution cause the processor to:
restart the rendering process in response to determining that an amount of time the 3D graphics processing unit takes to respond to commands before the rendering process terminated is less than a threshold.
12. The computer-readable storage device of claim 8 , wherein the computer readable instructions that upon execution cause the processor to restart the rendering process further comprise computer readable instructions that upon execution cause the processor to:
restart the rendering process in response to determining that an amount of available graphics memory controlled by the 3D graphics processing unit before the rendering process terminated is greater than a threshold.
13. The computer-readable storage device of claim 8 , wherein the computer readable instructions that upon execution cause the processor to restart the rendering process further comprise computer readable instructions that upon execution cause the processor to:
restart the rendering process and assign the rendering process to render on a second 3D graphics processing unit.
14. A method for recovering from a 3D graphics processing unit reset, comprising:
establishing, by virtualization platform, a communication channel between the virtualization platform and a virtual machine;
running, by the virtualization platform, a rendering process configured to render graphics for the virtual machine on a 3 D graphics processing unit;
detecting, by the virtualization platform, that the rendering process terminated;
restarting, by the virtualization platform, the rendering process;
sending, via the communication channel, a message indicating that the rendering process was restarted to the virtual machine;
establishing, by a virtual device driver executing within the virtual machine, a graphics buffer in a range of guest memory; and
sending, via the graphics buffer, 3D graphics data to the restarted rendering process.
15. The method of claim 14 , further comprising:
restarting the rendering process in response to determining that the virtual machine was not conducting a denial of service attack before the rendering process terminated.
16. The method of claim 14 , further comprising:
restarting the rendering process in response to determining that an amount of time the 3D graphics processing unit takes to respond to commands before the rendering process terminated is less than a threshold.
17. The method of claim 14 , further comprising:
restarting the rendering process in response to determining that an amount of available graphics memory controlled by the 3D graphics processing unit is greater than a threshold.
18. The method of claim 14 , further comprising:
restarting the rendering process and assign the rendering process to render on a second 3D graphics processing unit.
19. The method of claim 14 , further comprising:
restarting the rendering process in response to determining that an average latency of direct memory access transactions initiated by the virtual device driver before the rendering process terminated is less than a threshold.
20. The method of claim 14 , further comprising:
resending a frame of graphics data via the graphics buffer.