Memory power fault resilience in information handling systems
A controller of an information handling system may detect a power fault event for one or more of a plurality of memories configured to operate in a memory mirroring mode. The controller may deactivate the one or more of the plurality of memories by mapping the one or more of the plurality of memories out from usage without rebooting the information handling system based, at least in part, on the detection of the power fault event and the received notification.
1. A method, comprising:
detecting, by a controller of an information handling system, a power fault event for one or more of a plurality of memories controlled by the controller and configured in a memory mirroring mode with at least some information identical to information stored on another one or more of the plurality of memories; and
deactivating, by the controller, the one or more of the plurality of memories by mapping the one or more of the plurality of memories out from usage without rebooting the information handling system based, at least in part, on the detection of the power fault event and a received notification indicating that one or more of the plurality of memories of the information handling system are configured in the memory mirroring mode.
2. The method of claim 1 , wherein detecting, by the controller, a power fault event comprises detecting that a power good (PG) signal for the one or more of the plurality of memories has fallen below a predetermined threshold.
3. The method of claim 1 , wherein each of the plurality of memories comprises one or more dual in-line memory modules (DIMMs).
4. The method of claim 1 , further comprising: receiving, by the controller of the information handling system during a power-on self-test (POST) boot phase of the information handling system, a notification that each of the plurality of memories is configured in the memory mirroring mode.
5. The method of claim 1 , further comprising:
detecting, by the controller, a rising edge of a power good (PG) signal for at least one of the plurality of memories, during a power-on self-test (POST) boot phase during a reboot of the information handling system; and
initializing, by the controller and based, at least in part, on the detection of the rising edge of the PG signal, monitoring for runtime power fault events for the at least one of the plurality of memories.
6. The method of claim 5 , further comprising:
receiving, by the controller from a basic input output system (BIOS) of the information handling system, a message identifying at least one non-energized memory of the plurality of memories, during the POST boot phase,
wherein initializing, by the controller, monitoring for runtime power fault events for the at least one of the plurality of memories comprises initializing monitoring for runtime power fault events for each of the plurality of memories other than the at least one non-energized memory identified by the received message.
7. The method of claim 1 , further comprising:
detecting a power good (PG) signal for the one or more of the plurality of memories for which the power fault event was detected that meets a predetermined threshold for a predetermined period of time; and
reactivating the one or more of the plurality of memories for which the power fault event was detected based, at least in part, on the detected power good (PG) signal.
8. The method of claim 7 , wherein reactivating the one or more of the plurality of memories comprises instructing, by the controller, a basic input output system (BIOS) of the information handling system to map the one or more of the plurality of memories into an available memory space of the information handling system at a next reboot of the information handling system.
9. An information handling system, comprising: one or more central processing units (CPUs) including one or more processors; a controller; and a plurality of memories controlled by the controller and configured to operate in a memory mirroring mode, wherein the controller is configured to perform steps comprising: detecting a power fault event for one or more of the plurality of memories controlled by the controller and configured in a memory mirroring mode with at least some information identical to information stored on another one or more of the plurality of memories; and deactivating the one or more of the plurality of memories by mapping the one or more of the plurality of memories out from usage by the information handling system without rebooting the information handling system based, at least in part, on the detection of the power fault event and a received notification indicating that one or more of the plurality of memories of the information handling system are configured in the memory mirroring mode.
10. The information handling system of claim 9 , wherein detecting a power fault event comprises detecting that a power good (PG) signal for the one or more of the plurality of memories has fallen below a predetermined threshold.
11. The information handling system of claim 9 , wherein each of the plurality of memories comprises one or more dual in-line memory modules (DIMMs).
12. The information handling system of claim 9 , wherein the controller is further configured to perform steps comprising receiving a notification that each of the plurality of memories controlled by the controller is configured in a memory mirroring mode during a power-on self-test (POST) boot phase of the information handling system.
13. The information handling system of claim 9 , wherein the controller is configured to perform steps further comprising:
detecting a rising edge of a power good (PG) signal for at least one of the plurality of memories, during a power-on self-test (POST) boot phase during a reboot of the information handling system; and
initializing, by the controller and based, at least in part, on the detection of the rising edge of the PG signal, monitoring for runtime power fault events for the at least one of the plurality of memories.
14. The information handling system of claim 13 , wherein the controller is further configured to perform steps comprising:
receiving, from a basic input output system (BIOS) of the information handling system, a message identifying at least one non-energized memory of the plurality of memories, during the POST boot phase,
wherein initializing monitoring, by the controller, for runtime power fault events for the at least one of the plurality of memories comprises initializing monitoring for runtime power fault events for each of the plurality of memories other than the at least one non-energized memory identified by the received message.
15. The information handling system of claim 9 , wherein the controller is further configured to perform steps comprising:
detecting a power good (PG) signal for the one or more of the plurality of memories for which the power fault event was detected that meets a predetermined threshold for a predetermined period of time; and
reactivating the one or more of the plurality of memories for which the power fault event was detected based, at least in part, on the detected power good (PG) signal.
16. The information handling system of claim 15 , wherein reactivating the one or more of the plurality of memories comprises instructing a basic input output system (BIOS) of the information handling system to map the one or more of the plurality of memories into an available memory space of the information handling system at a next reboot of the information handling system.
17. A method, comprising:
detecting a power fault event for a memory of an information handling system;
detecting a power good (PG) signal for the memory that meets a predetermined threshold for a predetermined period of time; and
reactivating the memory based, at least in part, on the detected power good (PG) signal meeting the predetermined threshold for the predetermined period of time.
18. The method of claim 17 , further comprising deactivating the memory by mapping the memory out from usage by the information handling system without rebooting the information handling system while maintaining an energized state of the memory based, at least in part, on the detection of the power fault event.
19. The method of claim 17 , wherein reactivating the memory is performed after rebooting the information handling system and comprises enabling address range mirroring for the memory of the information handling system.
20. The method of claim 19 , wherein reactivating the memory further comprises:
determining that the memory has not encountered a power fault event for a predetermined time following enablement of address range mirroring for the memory; and
disabling address range mirroring for the memory of the information handling system.