Chapter 6. Management 307This setting is Enabled by default. When an NMI is issued by a critical eventthe BMC performs the system to reset for recovering the system. The BMClogs the reboot and additional error events in the SEL. Reboot on SPINTWhen any unrecoverable error condition occurs, the service processorinterrupt (SPINT) routine catches chipset register information of the machinecheck (MCK) error registers and stores it in the BMC NVRAM. Each maincomponent of the IBM eX4 is embedded by a different machine check errorregister that monitors the system operability and holds the state of thatcondition.This information is cleared after the system is recovered by a reboot. Howeverthe BMC starts the SPINT routine to store the information in its NVRAM area.This information is available, until the next MCK occurs.An MCK is reported in the BMC and is caused by a fatal situation that cannotbe handled by the chipset. Typically, an MCK is issued by an essentialcomponent of the IBM eXA4 chipset, particular components such as a DIMM,processor, or I/O linked devices. Ring Dump after SPINTThis setting is Disabled by default. By enabling the setting, the BMC catchesRing Dump information from the chipset, after an MCK occurred, to logunexpected error conditions. IBM Technical Support might request you toenable this setting.6.1.7 BMC firmware updateThe BMC can be flashed by a bootable image or by Linux or a Windows-basedupdate package. You may download updates from the IBM System x supportsite: BMC flash update (DOS bootable ISO image)http://www.ibm.com/support/docview.wss?uid=psg1MIGR-5073127 BMC flash update (Windows executable)http://www.ibm.com/support/docview.wss?uid=psg1MIGR-5073128 BMC flash update (Linux executable)http://www.ibm.com/support/docview.wss?uid=psg1MIGR-5073129Note: We recommend reporting a machine check error to IBM TechnicalSupport, if the information is not sufficient in the System x3850 M2 andx3950 M2 Problem Determination and Service Guide.
