• HIPER/Pervasive: DISRUPTIVE:  In response to recently reported security vulnerabilities, this firmware update is being released to address Common Vulnerabilities and Exposures issue number CVE-2018-3639. In addition, Operating System updates are required in conjunction with this FW level for CVE-2018-3639.
  

• HIPER/Pervasive:   A firmware change was made to address a rare case where a memory correctable error on POWER9 servers may result in an undetected corruption of data.
• A problem was fixed for Live Partition Mobility (LPM) to prevent an error in the hardware page translation table for a migrated page that could result in an invalid operation on the target system.  This is a rare timing problem with the logic used to invalidate an entry in the memory page translation table.
• A problem was fixed for a hung ethernet port on the service processor.  This hang prevents TCP/IP network traffic from the management console and the Advanced System Management Interface (ASMI) browsers.  It makes it appear as if the service processor is unresponsive and can be confused with a service processor in the stopped state.  An A/C power cycle would recover a hung ethernet adapter.
• A problem was fixed for partition hangs or aborts during a Live Partition Mobility (LPM) or Dynamic Platform Optimizer (DPO) operation.  This is a rare problem with a small timing window for it to occur in the hypervisor task dispatching.  The partition can be rebooted to recover from the problem.
• A problem was fixed for service processor static IP configurations failing after several minutes with SRC B1818B3E.  The IP address will not respond to pings in the ethernet adapter failed state.  This problem occurs any time a static IP configuration is enabled on the service processor.  Dynamic Host Control Protocol (DHCP) dynamic IPs can be used to provide the service processor network connections.  To recover from the problem, the other ethernet adapter (either eth0 or eth1) should be in the default DHCP configuration and allow the failing adapter to be reconfigured with a dynamic IP.
• A problem was fixed for the system going to ultra turbo mode after an On-Chip Controller (OCC) reset.  This could result in a power supply over current condition.  This problem can happen when the system is running a heavy workload and then a power mode change is requested or some error happens that takes the OCC into a reset.
• A problem was fixed for Workload Optimized Frequency (WOF) where parts may have been manufactured with bad IQ data that requires filtering to prevent WOF from being disabled.
• A problem was fixed for transactional memory that could result in a wrong answer for processes using it.  This is a rare problem requiring L2 cache failures that can affect the process determining correctly if a transaction has completed.
• A problem was fixed for a change in the IP address of the service processor causing the system On-Chip Controller (OCC) to go into Safe mode with SRC B18B2616 logged.  In Safe mode, the system is running with reduced performance and with fans running at high speed.  Normal performance can be restored concurrently by a reset/reload of the service processor using the ASMI soft reset option.  Without the fix, whenever the IP address of the service processor is changed, a soft reset of the service processor should be done to prevent OCC from going into Safe mode.
• A problem was fixed for the recovery for optical link failures in the PCIe expansion drawer with feature code #EMX0.  The recovery failure occurs when there are multiple PCIe link failures simultaneously with the result that the I/O drawers become unusable until the CEC is restarted.  The hypervisor will have xmfr entries with "Sw Cfg Op FAIL" identified.  With the fix, the errors will be isolated to the PCIe link and the I/O drawer will remain operational.
• A problem was fixed for a system aborting with SRC B700F105 logged  when starting a partition that the user had changed from P8 or P9 compatiblity mode to P7 compatibility mode.  This problem is intermittent and the partition in question had to have an immediate shutdown done prior to the change in compatibility mode for the problem to happen.  To prevent this problem when it is known that a compatibility mode is going to change to P7 mode, allow the partition to shut down normally before making the change.  If an immediate shut down of the partition is necessary and the compatibility mode has to be changed to P7, then the CEC should be powered off and then re-IPLed before making the change to prevent an unplanned outage of the system.
• A problem was fixed for a logical partition hang or unpredictable behavior due to lost interrupts with BCxxE504 logged when memory is relocated by the hypervisor because of predictive memory failures.  This problem is not frequent because it requires memory failing and the recovery action of relocating memory away from the failing DIMMs being taken.  To circumvent this failure, if memory failure has occurred, the system may be re-IPLed to allow normal memory allocations from the available memory, so the partitions do not have to run on relocated memory.
  
• A problem was fixed for a failure in DDR4 RCD (Register Clock Driver) memory initialization that causes half of the DIMM memory to be unusable after an IPL.  This is an intermittent problem where the memory can sometimes be recovered by doing another IPL.  The error is not a hardware problem with the DIMM but it is an error in the initialization sequence needed get the DIMM ready for normal operations.
  

System firmware changes that affect certain systems

• DEFERRED:   On a system with only a single processor core configured, a problem was fixed for poor I/O performance.  This problem may be circumvented by configuring a second processor core.  This additional processor core does not have to be used by the partition.
• On systems that are not managed by a HMC, a problem was fixed to enable FSP call home.  The problem always happens when the service processor attempts to call home an error log as it will fail.
• A problem was fixed for Dynamic Power Saver Mode on a system with low-CPU utilization that had reduced performance unexpectedly.  This only happens for workloads that are restricted to a single core or using just a single core because of low-CPU utilization.  This problem can be circumvented by running the system in maximum performance mode by using ASMI to enable Fixed Maximum Frequency mode.
  

• GA Level