Lenovo Storage Adapter Firmware



Supported Adapters:
	ThinkSystem RAID 9350-8i 2GB Flash PCIe 12Gb Adapter	
	ThinkSystem RAID 9350-16i 4GB Flash PCIe 12Gb Adapter
    ThinkSystem RAID 9350-16i 4GB Flash PCIe 12Gb Internal Adapter
    ThinkSystem RAID 9350-8i 2GB Flash PCIe 12Gb Internal Adapter	
	ThinkSystem RAID 5350-8i PCIe 12Gb Adapter
    ThinkSystem RAID 5350-8i PCIe 12Gb Internal Adapter
	ThinkSystem 4350-8i SAS/SATA 12Gb HBA
	ThinkSystem 4350-16i SAS/SATA 12Gb HBA

Version 5.91
-------------------------------------
    FW Version: 5.91

Fixed issue:
-Fixed an issue where the board was bricked when the ROM toggle command was given by the application
-Fixed an issue where a certain sensor was not found.
-Fixed an issue wherein the temperature sensor location value is reported incorrectly in the tools for some HPE
Gen10 boards.
-Fixed an issue where volumes are no longer present when a Reset to factory defaults is done when the
preserve volume is enabled
-Fixed an issue where excessive Smart Carrier failure messages are being sent to the controller log.
-Fixed an issue where the PLDM event consumer was missing events from the previous boot.
-Fixed an issue where the consistency check completion date is reported wrong.
-Fixed an issue wherein the ILA scratchpad size is too small.
-Fixed an issue wherein the controller had a lock-up due to an LOS during a sequential write traffic
-Fixed an issue where an incorrect treatment of the SMP Discover response happened with the ATTACHED
SATA DEVICE set and ATTACHED SAS DEVICE TYPE bits set to 0.
-Fixed an issue where firmware incorrectly reported unrecoverable media errors.
-Fixed an issue where a spare rebuild is aborted if a Foreign SED is added as a replacement drive.
-Fixed an issue where PLDM event consumer was missing events from the previous boot.
-Fixed an issue where Managed SED secured logical drive status goes to SED DATA LOCKED state after
inserting a Foreign SED.
-Fixed an issue where the degraded/failed non-fault tolerant Managed SED secured logical drive is moved to OK
state on inserting a Foreign SED.
-Fixed an issue where a Degraded or Failed Managed SED secured logical drive is moved to an incorrect state
when inserting a Foreign SED.
-Fixed an issue where Rapid Parity Initialization (RPI) is resuming on SED DATA LOCKED logical drive after
reboot.
-Fixed an issue where the consistency check completion date is incorrect
-Fixed an issue with logical drive status showing FAILED(MEx) instead of LOOSE_CABLE.
-Fixed the clear controller configuration command failure
-Fixed an issue of rebuild getting reiterated continuously and not progressing
-Fixed an issue where rebuilding did not start on Managed SED secured logical drive when the adapter password
is enabled.
-Fixed an issue where rebuild was completed only for the first logical drive in a Managed SED secured array.
-Fixed an issue where rebuild starts on a failed Managed SED secured fault-tolerant logical drive when a Foreign
SED is imported.
-Fixed an issue of Over Provisioning Optimization (OPO) being initiated on a dedicated spare SSD drive.
-Fixed an issue where the spare rebuild stops when a Foreign SED is inserted
-Fixed an issue where the logical drive state changed to FAILED(MEx) instead of FAILED.
-Fixed an issue where the spare rebuilding is not happening when a global dedicated spare is assigned for
multiple arrays.
-Fixed an issue where the required reason for controller reboot is not getting updated for the sanitize policy
change.
-Fixed an issue where the spare rebuild is not started on the degraded logical drive, even after the global spare is
released from the Heal Array operation
-Fixed an issue where the logical drive moved to the FAILED state upon hot adding the spare replaced drive.
-Fixed an issue where the SED logical drive is not listed at OS level after migration and import from another
server.
Fixed an issue where the controller reports sanitizing complete on a hot-removed and hot-plugged back in HDD
that happens while sanitize is being performed.
Fixed an issue where the Foreign SED remains foreign after changing master key to match Foreign SED.
-Fixed an issue where a reboot happens while a Managed SED logical drive is being transformed and after
rebooting, the cache was dirty during the reboot and the logical drive came up in an Abnormal Volume state.
-Fixed an issue where a reboot during a logical drive move operation can result in unused SED drives not getting
restored to Original Factory State (OFS).
-Fixed an issue of the Managed SED adapter password internal lockout wait timeout being reset on system
reboot.
-Fixed an issue where firmware is processing additional unlock adapter password request even if the controller is
already unlocked.
-Fixed an issue where deleting a non-existent secure logical drive results in a TLB Exception lockup
-Fixed an issue of power surge when too many SATA drives are spinning up simultaneously
-Fixed an issue where the firmware debug command is displaying secret data.
-Fixed an issue where the failed SED drive is found during drive spin-up
-Fixed an issue where the datastore is updated, when SED is moved to “Otherwise Owned”.
-Fixed an issue of a certain sensor not being found.
-Fixed an issue where the controller looks to be unresponsive as the firmware is waiting on the adapter password
and a data drive in a secured volume is removed and replaced with an HBA/unconfigured locked SED.
-Fixed an issue where a hot-added Managed SED is reported to the host before it is fully ready.
-Fixed an issue where on reboot after panic shutdown, Ownership or Revert of the drive is unsuccessful
-Fixed an issue where on reboot after panic shutdown, non-secure logical drive creation fails.
-Fixed an issue when a replacement SED in a secured logical drive is marked as failed during hot-plug or bootup
when the Managed SED is disabled
-Fixed an issue where logical drives are no longer present after a controller Reset to factory defaults when the
preserve logical drive is enabled.
-Fixed an issue where the OS boot up might be slow after a cold boot because a logical drive is stuck in the
Rebuilding state. This logical drive has an active spare locked SED (configured as a Dedicated Spare) and the
previous failed drive must be hot-plugged back into the logical drive in Passive SED configuration
-Fixed an issue where Volume CREATE requests fail if a deferred binding string for a data drive resource was
passed in the request payload.
-Fixed an issue where an UpdateComponent request for a drive will still be accepted with an invalid
ComponentImageSize.
-Fixed an issue where DurableName associated with each LUN on a Multi-Actuator drive is published with
incorrect values.
-Fixed an issue where the WriteCacheDataLoss event was not sent when expected
-Fixed an issue where an incorrect extended error message was sent when attempting a SecureErase ACTION
on a SED that is not in its OFS.
-Fixed an issue where the updateTime field in the GetPDRRepositoryInfo command response was not being
updated correctly.
-Fixed an issue where the BatteryMissing event was not being sent when the battery was removed prior to boot
-Fixed an issue where duplicate entries for an expander were seen in the Type 5 downstream device inventory
-Fixed an issue where a DriveOffline Redfish alert is not sent when a Drive resource begins a sanitize operation
-Fixed an issue where incorrect CapabilitiesDuringUpdate flags were set for UBM PICs which do not support
firmware updates
-Fixed an issue where an unflashable SMP PSOC device was appearing in the Type 5 downstream device
inventory.
-Fixed an issue where hot-removed SEPs could prevent other remaining SEPs from being included in the Type 5
downstream device inventory
-Fixed an issue where a VolumeOffline Redfish alert could be sent with an incorrect severity in certain conditions.
-Fixed an issue where a controller firmware update through PLDM fails if the final requested image segment is
smaller than the spec-defined minimum transfer size.

Version 5.33
-------------------------------------
    FW Version: 5.33
	
Fixed issue:
-Added support for the Lenovo NEW34 SAS backplane wherein the Lenovo SKUs are set to use the auto backplane detection and all SKUs have internal connection only. Each internal connector allocates one element out of an internal enclosure array which is shared by the SAS enclosure and the SGPIO enclosure. The SAS enclosure always initialize before the SGPIO enclosure. When the Lenovo NEW34 SAS backplane is connected to a Lenovo HBA, the backplane type detected is SGPIO. As a result, the firmware will proceed and set up the
SGPIO enclosure. Upon the SGPIO enclosure initialization, the firmware makes sure that it does not overwrite the SAS enclosure data.
-Added support for reporting the previous cache read/write setting to the PLDM when the only_read cache setting happens. A sense feature page 0x10 in BMIC 0x61 and a save user cache setting in BMIC 0xC1 were added.
-Added support for the firmware to hide the NVMe drive attached to the Lenovo UBM backplane. The Lenovo "Anybay" backplane can be set up to route the SAS/SATA drive to the Lenovo HBA and the NVMe drive to the PCIe. As a result, the firmware does not need to expose any NVMe drive found, to the UFEI/BIOS and OS driver. It will not mark it as a failed device either.
-Added SEEPROM and manufacturing support for the Lenovo card 5350-8i-INTR. As a result, a new SEEPROM file and manufacturing data file were added.
-Added FUP support for the Lenovo card 5350-8i INTR. As a result, new SKU PCI IDs were added in the firmware build script.
-Added support to enable random logical drive deletion without a backup power source on the controllers supporting data preservation. Firmware disk-based transformation limitation of transformation backup memory has been increased from 8192 sectors (4 MiB) to 28672 sectors (14 MiB). Since the maximum safe stripe size is 12 MiB for an 8GB controller, it is sufficient to increase the RIS area expansion size to 14 MiB.
-Fixed an issue where the firmware does not abort all pending requests when the device LUN reset happens.
-Fixed an issue where the PBSI multi-lun field remains unset for multi-lun drives.
-Fixed an issue wherein a ~8% performance drop occurred on HBA mode for 4K/8K random read with the IO queue depth of 64 or higher after upgrading the firmware version from v1.32 to v4.11.
-Fixed an issue where the RAID Stack continuously retries a command to the SEP device that was completed with the TASK SET FULL status.
-Fixed an issue where the range locked SEDs of the failed SED qualification volume are not coming to a good state after the volume is deleted.
-Fixed an issue where the encrypted single drive RAID 0 logical drive was failed, replaced and re-enabled from the previous boot. But after the reboot, the volume state of the same logical drive was marked as NEED_CRYPTO_KEY.
-Fixed an issue where the cache status shows 'temporary disabled' when the battery is removed.
-Fixed an issue where a wrong location is displayed for the temperature sensor 0.
-Fixed an issue where the pointers in PLDM commands were being accessed before being initialized.
-Fixed an issue where a data loss occurred in the encrypted RAID 50/60 logical drive when the logical drive failed and healed because of the drives in the hot removal/plug.
-Fixed an issue where duplicated drive entries were listed in the expander topology.
-Fixed an issue where an SED volume was marked as failed during the capacity expansion transformation.
-Fixed an issue where the function of creating a logical drive failed with no backup power source and Maxcache present.
-Fixed an issue where the controller hung when a heavy IO is run during a rebuild on any logical volume except RAID6.
-Fixed an issue where the volume state change is not correct when marking a drive as a dead drive.
-Fixed an issue where the slot number is showing as 'unknown' for the directly attached drives.
-Fixed an issue where the subsequent transformation was suspended after the system rebooted unsafely during a transformation.
-Fixed an issue where the persistent event logs were not captured through the debug log.
-Fixed an issue where the filesystem or application read incorrect data during random IO on a degraded RAID5 unit.
-Fixed an issue where the drive slot number is showing incorrectly for a failed drive.
-Fixed an issue where the filesystem or application read incorrect data on a degraded parity volume
-Fixed an issue where the filesystem or application read incorrect data on a degraded parity SSD volume
-Fixed an issue where the firmware was unable to update the parameters from the SmartIOC_2100-8i.cfg to the SmartIOC_2100-8i board.
-Fixed an issue where the Logical Volume went to FAILED state on drive(s) failure when there were enough spare drives in the array.
-Fixed an issue where the MCTP Discovery Notify cannot be configured
-Fixed an issue where the cold-boot during expansion caused an abnormal volume state.
-Fixed an issue with a UBM backplane where not all the drives were being discovered.
-Fixed an issue where the product ID of an enclosure was not correctly displayed
-Fixed an issue where the Predictive Fail LED did not blink during the predictive spare rebuild.
-Fixed an issue where the RPI was started on volume with SEDs and then the controller was rebooted despite the RPI start happening while the volume was locked.
-Fixed an issue where the controller hung during the BMIC clear config when one of the drives was missing and the missing drive index was assigned from a volume that did not finish RPI.
-Fixed an issue where the Long SATA SSD TRIM caused a hang.
-Fixed an issue where the PF LED blinked when there was a drive failure during an ongoing PSR.
-Fixed an issue where the UBM3 failed to flash.
-Fixed a potential UBM backplane flash issue that may occur if the image start index was not aligned with the start sector.
-Fixed an issue where the RTC timestamp was not being sent to the SES based storage enclosure SEPs internal to the enclosure.
-Fixed an issue where 2 UBM backplanes connected to one connector and one UBM backplane incorrectly identified itself as not on a bifurcated cable.
-Fixed an issue where the PLDM event needed to be sent when the controller password was entered and no encrypted volumes were present

Version 5.00
-------------------------------------
    FW Version: 5.00
Fixed issue:
-Fixed an issue where the Lenovo CFF 9350-16i has an incorrect Phy Mapping
-Fixed an issue where the direct attached drive without management does not support the zero-based bay
numbering scheme.
-Fixed an issue where the firmware will assign a bay number to all the slots on a SGPIO backplane detected
-Fixed an issue where the 0x1789 driver health error code is displayed as healthy when the controller is in an
abnormal volume state wherein all the logical drives including the healthy ones are offline.
-Added support for PLDM Type 6 Basic Volume Creation
-Added support for RetrieveCustomResponseParameters command to retrieve the NewResourceID.
-Added support fpr PLDM Type 6 Port.ActiveWidth
-Added support for Storage alerts for WriteCacheProtected. WriteCacheTemporarilyDegraded,
WriteCacheDegraded and WriteCacheDataLoss. When the storage WriteCache changes state, the
corresponding Redfishalert event is generated; of which, the MC can poll or asynchronously receive the event.
-Added support for PLDM Type2 Numeric Sensor for Supercap Temp.
-Added logic to add a numeric sensor PDR representing a supercap temperature sensor to the numeric sensor
PDR repo table if a supercap backup power source is present on the controller.
-Added Privilege Operation support.
-For certain supported platforms, support for the following RDE custom headers has been added:
 + PLDM-RDE-MCHP-PRIVILEGE-TOKEN
 + PLDM-RDE-MCHP-REVOKE-PRIVILEGE-TOKEN
Note: Volume CREATE and DELETE operations may be rejected with ERROR_NOT_ALLOWED under
certain conditions on platforms where these custom headers are supported.
-Added support for PLDM Type 6 Delete Volume.
-Fixed issue wheere With COLD_REBOOT flag set, ApplyComplete is 0 instead of expected 1.
-Fixed issue where when Volume.InitializeMethod=Foreground (RPI), the Health is set to Warning instead of Ok.
-Fixed an issue where there was an incorrect controller port speed in the Redfish Port schema wherein the
response using an RDE client showed different max and current speeds than expected.
-Fixed an issue where the PLDM Type 6 StorageController resource odata.id is empty and the empty odata.id is
published on an RDE GET operation for a StorageController resource
-Fixed an issue where Volume alerts were missing after a reboot.
-Fixed an issue where Volume READ publishes InitializeMethod = Background for a Volume queued for RPI
wherein creating two volumes on the same array/set of drives with an InitializeMethod of Foreground, a READ
operation on the second volume will report an InitializeMethod of Background.
	

Version 4.72
-------------------------------------
    FW Version: 4.72
Fix issue:
-Fixed an issue where LUN reset for SEP device completes with an incorrect LUN status to the host.
-Fixed an issue where drives are still discovered when the firmware reports a "Memory Self-Test Error" message.
-Fixed issues in SED qualification and in recovery of SED qualification failed drives.
-Fixed a potential LUN reset occurring on the controller's virtual SEP device due to a timeout for the Notification On Event (0x40) command.
-Fixed an issue where the SED HBA drive security status does not update after changes are made.
-Fixed an issue where WWID is incorrectly displayed in UART
-Fixed an issue for error "12G Unknown eeprom v2 addr".
-Fixed an issue where Luxor FW failed to configure the Purley (Pegasus) SGPIO backplane properly.
-Fixed an intermittent warning message of "Logical drive is degraded due to loss of redundant path" being displayed in the SSA GUI utility when a logical drive is in a failed state and in a dual domain configuration.
-Fixed an issue where the Fault LED is not turned on when the RAID Stack failed a bad HBA drive connected to the expander during device discovery or hot-plug.
-Fixed an issue where the LUN reset from the host to the same drive continuously occurs every 60 seconds after injecting crash on the Natasha Peabody expander while running IOmeter test on the 4 drives RAID-0 volume with the smart/AIO path enabled.
-Fixed a problem where the host WRITE I/O failed on the SmartCache logical drive with UREs present on the primary logical drive.
-Fixed an issue where hot spare auto rebuild does not start when two members of RAID 10 fail simultaneously
-Fixed an issue where marking a URE block on a drive is not successful.
-Fixed an issue where transformation is not suspended even after the expand disable code is set.
-Fixed a performance drop on non-parity HDD volumes when coalescing was on.
-Fixed an issue with the Fault LED wherein it is not lit on Predictive fail drive when volume is a mix of SED and non-SED drives.
-Fixed an issue with the Fault/Locate LED setting wherein if multiple configured drives are hot removed from a SES enclosure and then hot added, the Fault LED is turned on for a few hot added drives which indicates a drive failure for a few seconds.
-Fixed an issue wherein the enclosure fails to receive the SES control from the controller when drives are partially populated.
-Fixed an issue wherein Luxor sends SesCtrl commands to the wrong bay number after an enclosure power cycle
-Fixed an issue wherein the Physical PHY to logical PHY mapping was wrong for the Lenovo CFF 9350-16i-INTR controller.
-Fixed an issue wherein external SAS/SATA cable errors never got cleared.
-Fixed an issue wherein the cable data is not being read after it is added.
-Fixed an issue wherein the SmartCache will not exit the flushing state.
-Fixed an issue wherein the PHY rates cannot be adjusted when the connector limit is adjusted.
-Fixed an issue wherein a performance drop was introduced as a result of a recent code change.
-Fixed an issue wherein the URE is not logged for configured HDD drives after a reboot.
-Fixed an issue wherein the Diskname is not showing for drives after the deletion of the Logical drive and RIS data present on the SEDs.
-Fixed an issue where the firmware can always report change pending if the discovery protocol option is auto detect.
-Fixed an issue wherein the file system may read incorrect data from a degraded parity logical drive.

Version 3.82 - initial Release
-------------------------------------
    FW Version: 3.82