Preserving error context during a reboot of a computing device
To preserve error context during a reboot of a computing device, firmware within the computing device can be configured to implement a method that includes determining where the error context is stored in volatile memory. The method can also include identifying a plurality of recorder regions in non-volatile memory that have been assigned to store the error context. The plurality of recorder regions can be disaggregated across a plurality of distinct non-volatile memory regions. The method can also include flushing the error context from a plurality of different volatile memory locations to the plurality of recorder regions in response to detecting a trigger. The flushing can occur prior to the reboot of the computing device. The method can also include restoring at least some of the error context to the volatile memory after the computing device has been rebooted.
1. A method for preserving error context during a reboot of a computing device, the method being implemented by firmware within the computing device, the method comprising:
determining where the error context is stored in volatile memory;
identifying a plurality of recorder regions in non-volatile memory that has been assigned to store the error context, wherein the plurality of recorder regions is disaggregated across a plurality of distinct non-volatile memory regions;
communicating an indication of a recorder region of the plurality of recorder regions to an operating system of the computing device;
causing the error context from the volatile memory to be saved to the plurality of recorder regions in the non-volatile memory in response to detecting a trigger, wherein the error context is saved to the plurality of recorder regions prior to the reboot of the computing device; and
restoring at least some of the error context to the volatile memory after reboot of the computing device.
2. The method of claim 1 , wherein the plurality of recorder regions comprises at least one of:
a first recorder region in a firmware region of the non-volatile memory;
a second recorder region in a non-volatile memory partition that is created during a boot process of the computing device; or
a third recorder region in an auxiliary service processor.
3. The method of claim 1 , wherein the error context includes operating system context for the operating system within the volatile memory.
4. The method of claim 1 ,
wherein a location of the error context in the volatile memory is assigned during a boot process of the computing device.
5. The method of claim 4 , further comprising receiving a notification from the operating system about at least one additional region of the volatile memory that the operating system has assigned to store the error context.
6. The method of claim 1 , wherein the at least some of the error context is included in one or more registers within one or more hardware devices that are part of the computing device.
7. The method of claim 6 , wherein determining where the error context is stored in the volatile memory comprises determining that the one or more registers have been assigned to store the error context when the one or more hardware devices are initialized during a boot process of the computing device.
8. The method of claim 1 , wherein:
the firmware comprises system management mode (SMM) code that is executed by a processor of the computing device when the processor enters a SMM; and
causing the error context to be saved to the plurality of recorder regions comprises executing at least a part of the SMM code.
9. The method of claim 1 , wherein detecting the trigger comprises detecting a system management interrupt (SMI) signal.
10. The method of claim 9 , wherein:
the computing device is part of a cloud computing system that comprises a management entity;
the computing device comprises an auxiliary service processor; and
the auxiliary service processor causes the SMI signal to be generated in response to a command from the management entity.
11. The method of claim 1 , further comprising causing the at least some of the error context to be transmitted to a remote location for analysis.
12. A computing device, comprising:
one or more processors; and
firmware that is executable by the one or more processors to:
determine where error context is stored in volatile memory;
identify a plurality of recorder regions in non-volatile memory that has been assigned to store the error context, wherein the plurality of recorder regions is disaggregated across a plurality of distinct non-volatile memory regions;
communicate an indication of a recorder region of the plurality of recorder regions to an operating system of the computing device;
cause the error context from the volatile memory to be saved to the plurality of recorder regions in the volatile memory in response to detecting a trigger, wherein the error context is saved to the plurality of recorder regions prior to the reboot of the computing device; and
restore at least some of the error context to the volatile memory after reboot of the computing device.
13. The computing device of claim 12 , wherein the plurality of recorder regions comprises at least two of:
a first recorder region in a firmware region of the non-volatile memory;
a second recorder region in a non-volatile memory partition that is created during a boot process of the computing device; or
a third recorder region in an auxiliary service processor.
14. The computing device of claim 12 , wherein at least some of the error context is included in at least one of:
an operating system context within the volatile memory; or
one or more hardware registers within one or more hardware devices that are part of the computing device.
15. The computing device of claim 12 , wherein:
detecting the trigger comprises detecting a system management interrupt (SMI) signal;
the computing device is part of a cloud computing system that comprises a management entity;
the computing device further comprises an auxiliary service processor; and
the auxiliary service processor causes the SMI signal to be generated in response to a command from the management entity.
16. The computing device of claim 12 , wherein the firmware is additionally executable by the one or more processors to cause the at least some of the error context to be transmitted to a remote location for analysis.
17. A cloud computing system, comprising:
a management entity; and
a plurality of servers, each server among the plurality of servers comprising one or more processors, an auxiliary service processor, and firmware that is executable by the one or more processors to:
identify a plurality of recorder regions in non-volatile memory that has been assigned to store the error context, wherein the plurality of recorder regions is disaggregated across a plurality of distinct non-volatile memory regions;
communicate an indication of a recorder region of the plurality of recorder regions to an operating system of the computing device
cause the error context from the volatile memory to be saved to the plurality of recorder regions in the non-volatile memory in response to detecting a trigger, wherein the error context is saved to the plurality of recorder regions prior to the reboot of the computing device; and
transmit at least some of the error context to a remote location for analysis.
18. The cloud computing system of claim 17 , wherein the plurality of recorder regions comprises:
a first recorder region in a firmware region of the non-volatile memory;
a second recorder region in a non-volatile memory partition that is created during a boot process of a computing device; and
a third recorder region in an auxiliary service processor.
19. The cloud computing system of claim 17 ,
wherein the error context includes operating system context for the operating system within the volatile memory; and
wherein the at least some of the error context is included in one or more registers within one or more hardware devices that are part of the computing device.
20. The cloud computing system of claim 17 , wherein the firmware is implemented in a server and is additionally executable by the one or more general purpose processors to restore at least some of the error context to volatile memory after reboot of the server.