Printable version

Drivers & software

HP rx8640/rx7640 and Server Expansion Unit (SEU) Firmware

By downloading, you agree to the terms and conditions of the Hewlett Packard Enterprise Software License Agreement.
Note:  Some software requires a valid warranty, current Hewlett Packard Enterprise support contract, or a license fee.

Type: Firmware (Entitlement Required) - System
Version: 4.2(MPI Initiated)(18 Oct 2010)
Operating System(s):
HP-UX 11.23 | View all
File name: m_ipf_sx2000_4_2.tar.gz (8.7 MB)
Firmware release for HP rx8640, rx7640, and Server Expansion Unit (SEU)

End User License Agreements:
HPE Software License Agreement v1


HP rx8640/rx7640 and Server Expansion Unit (SEU) Firmware Release

TITLE:HP rx8640/rx7640 and Server Expansion Unit (SEU) Firmware Release

VERSION: rx8640/rx7640 and SEU (AB301A) Firmware Version 4.2

Programmable Hardware:

  System Backplane GPM   	:   001.002.000
  System Backplane EMMUX 	:   001.000.000
  IO Backplane IO_LPM-0  	:   001.004.000
  IO Backplane IO_LPM-1  	:   001.004.000
  PCIe IO Backplane IO_LPM-0	:   002.003.000
  PCIe IO Backplane IO_LPM-1	:   002.003.000
  Core IO CIO            	:   001.002.000
  Cell_LPM               	:   001.002.000
  Cell_JTAG              	:   001.002.000
  Cell_PDH               	:   001.005.001
Firmware:
  Core IO MP             	: B.004.002.003
  Core IO ED			:   002.011.000
  Core IO SCSI FW             	:   001.003.035.070
  Core IO SCSI EFI Driver	:   001.005.004.000
  Cell PDHC              	: B.023.003.040
  Cell IPF_FW            	:   009.066.000

DESCRIPTION:
This firmware release supports the HP rx8640/rx7640 servers with the following processors:

IPF Montvale CPUs:

  • 1.6GHz 24MB cache dual-core Intel (R) Itanium (R) processor 9000 series
  • 1.6GHz 18MB cache dual-core Intel (R) Itanium (R) processor 9000 series
  • 1.4GHz 12MB cache dual-core Intel (R) Itanium (R) processor 9000 series

IPF Montecito CPUs:

  • 1.6GHz 24MB cache dual-core Intel (R) Itanium (R) processor
  • 1.6GHz 18MB cache dual-core Intel (R) Itanium (R) processor
  • 1.4GHz 12MB cache dual-core Intel (R) Itanium (R) processor

IPF Mad9M CPUs:

  • 1.6 GHz 6MB L3 cache single-core Intel (R) Itanium (R) processors

UPDATE RECOMMENDATION: Recommended

HP recommends systems used in Mission Critical environments update to this revision of firmware during the next system maintenance period in order to improve system reliability.

[ ] Critical
    [ ] Panic, [ ] Hang, [ ] Abort, [ ] Corruption, [ ] Memory Leak,
    [ ] Other _______________
    [ ] Hardware Enablement, [ ] Software Enablement
[ ] Required
[X] Recommended
[ ] Optional
    [ ] Hardware Enablement, [ ] Software Enablement [ ] non-critical
[ ] Initial Customer Release

SUPERSEDES: 4.1

PRODUCT MODEL(S):

  • HP Integrity rx7640 8-socket Enterprise Server
  • HP Integrity rx8640 16-socket Enterprise Server
  • HP Integrity rx8640 16 Server Expansion Unit (SEU)

OPERATING SYSTEMS:

For dual-core Intel (R) Itanium (R) processor, Montvale:

  • HP-UX 11i v2 OE Update Release for June 2007 or later; Support for vPars A.04.04
  • HP-UX 11i v3 OE Update Release for Sept 2007 or later; Support for vPars A.05.02
  • Microsoft Windows Server 2003 64-Bit Edition
  • Microsoft Windows Server 2008 Itanium
  • Microsoft Windows Server 2008 Enterprise and Data Center
  • Microsoft Windows Server 2008 SP2 Enterprise and Data Center
  • Microsoft Windows Server 2008 R2 Enterprise and Data Center
  • OpenVMS 8.3-1H1
  • OpenVMS 8.4
  • SUSE Linux Enterprise Server 10 Service Pack 1

For dual-core Intel (R) Itanium (R) processor, Montecito:

  • HP-UX 11i v2 OE Update Release for Sept 2006 or later; Support for vPars A.04.02
  • HP-UX 11i v3; Support for vPars A.05.01
  • Microsoft Windows Server 2003 64-Bit Edition
  • Microsoft Windows Server 2008 Itanium
  • Microsoft Windows Server 2008 Enterprise and Data Center
  • Microsoft Windows Server 2008 SP2 Enterprise and Data Center
  • Microsoft Windows Server 2008 R2 Enterprise and Data Center
  • OpenVMS 8.3
  • RedHat Enterprise Linux 4 Update 4
  • RedHat Enterprise Linux 5 Update 5
  • SUSE Linux Enterprise Server 10

For single-core Intel (R) Itanium (R) processor, Mad9M:

  • HP-UX 11i v2 OE Update Release for March 2006 or later; Support for vPars A.04.01
  • HP-UX 11i v3; Support for vPars A.05.01
  • Microsoft Windows Server 2003 64-Bit Edition
  • Microsoft Windows Server 2008 Itanium
  • Microsoft Windows Server 2008 Enterprise and Data Center
  • Microsoft Windows Server 2008 SP2 Enterprise and Data Center
  • Microsoft Windows Server 2008 R2 Enterprise and Data Center

LANGUAGES:
International English

ENHANCEMENTS:

  • In webGUI, Partition names are printed along with Partition numbers.
  • MP webGUI, vKVM and vMedia features now work in a NAT environment
    NOTE: A new option for the SA command has been added -
    "U - Hostname in Web URLs"
    This has to be set to "Enabled" for the MP to work in a NAT environment
    This has to be set to "Disabled" for the MP to work in a non-NAT environment
  • Numerous improvements have been made to the MP WebGUI help pages.
  • Improvements have been made to MP user administration.

FIXES:

Cell IPF FW 009.066.000 Contains the Following Fixes:

  • Certain cell/IO combinations do not boot Windows (STOP CODE A5). All supported cell/IO combinations should now boot.
  • In rare cases, an IPF vPar panic could occur when another vPar was being rebooted. This is now fixed.
  • With Windows, cab/bay/chassis is always displayed as 0/0/0 for embedded device in IO chassis number one. The correct path is now displayed.
  • Previously, not all LUNs can be seen from EFI in some large SAN configurations. All LUNs are now visible.

Core IO MP Firmware B.004.002.003 Contains the Following Fixes:

  • MP IP addresses ending in .13 cause failures when using the iLO GUI. These addresses now work.
  • The MP telnet server now correctly handles the Telnet NOP.
  • The MP is unreachable from 192.168.0.0 subnet when netmask is other than 255.255.255.0. This now works.
  • Issues could occur when launching vMedia from 2 different browser windows on the same client machine. This has been corrected.
  • After recovery from Master MP failover state, the LDAP configuration could be lost. This has been fixed.
  • CM inactivity timeout disabled by repeated MP login attempts no longer occurs.
  • Local user web sessions were not closed on a Security Configuration reset. These are now closed during this operation.
  • Deleting a user now closes all connections previously opened by the user.
  • Users can no longer login to iLO with blank/invalid credentials when LDAP is enabled.
  • Anonymous logon is now enabled in Microsoft Active Directory, enabling LDAP-lite to work.

PREREQUISITES:
None

INSTALLATION INSTRUCTIONS:

Please review all instructions and the Hewlett-Packard support terms and conditions for precautions, scope of license, restrictions, and limitation of liability and warranties before installing this patch. It is important to read and understand these instructions completely before beginning. This can determine the success of completing the firmware update.

Unpacking the downloaded file (PC or HPUX and Checksum Values):

Firmware files may be obtained from several locations. When downloading firmware files, all files should be stored in the same location (path). This allows a firmware update of all entities during one firmware update session.

Note: Always download the files to your PC in Binary Mode.

Files may be located via a web browser by:

  1. Accessing the URL http://www.hp.com
  2. Then choose "Software & driver downloads"
  3. Enter the product number (Example: rx8640)
  4. Then choose "Cross operating system (BIOS, Firmware, Diagnostics, etc.)"
  5. Select the package from the list to begin download

The following tools will be required for firmware updates:

  • Network console connection to a MP console port.
  • Laptop and 9 pin RS-232 serial cable if network console is not available.
  • Windows XP or HP-UX FTP server with firmware file from site listed above.

Downloading using a Windows FTP Server:

Note: The Microsoft Windows environment which the file is going to be downloaded to must have WinZip installed to un-package the m_ipf_sx2000_4_2.tar.gz file.

  1. FTP m_ipf_sx2000_4_2.tar.gz to an empty folder.
  2. Open a Windows Explorer window, select the drive and folder that m_ipf_sx2000_4_2.tar.gz was downloaded to.
  3. Double click on the m_ipf_sx2000_4_2.tar.gz file to start the WinZip program and the un-packaging process.
  4. The WinZip program will ask if the file should be decompressed into an empty folder and opened. Click on the "Yes" button.
  5. The WinZip program will start and display all of the files within the m_ipf_sx2000_4_2.tar.gz file. Click on the "Extract" button.
  6. WinZip will ask for the folder that the files should be extracted to. Using the Folders/Drives portion of the window select the desired location and then click on the "Extract" button.
  7. Using the Windows Explorer window, select the drive and folder where m_ipf_sx2000_4_2.tar.gz was downloaded.
  8. Individually "Right Click" on each of the files located in this folder and select "Properties". Verify that each file has the following properties:
          hm_ipf.009.066.000.frm
               Size: 12,584,000 bytes

          hm_pdhc.023.003.040.frm
               Size: 806,112 bytes

	  m_ap_lpm_0.002.003.000.frm
	       Size: 14,67,601 bytes

	  m_ap_lpm_1.002.003.000.frm
	       Size: 14,67,601 bytes

          m_cell_jtag.001.002.000.frm
               Size: 836,134 bytes

          m_cell_lpm.001.002.000.frm
               Size: 1,467,263 bytes

          m_cell_pdh.001.005.001.frm
               Size: 5,246,318 bytes

          m_cio.001.002.000.frm
               Size: 827,778 bytes

          m_ed.002.011.000.frm
               Size: 476,568 bytes

          m_emmux.001.000.000.frm
               Size: 821,051 bytes

          m_gpm.001.002.000.frm
               Size: 1,463,679 bytes

          m_io_lpm_0.001.004.000.frm
               Size: 1,463,679 bytes

          m_io_lpm_1.001.004.000.frm
               Size: 1,463,679 bytes

	  m_mp.004.002.003.frm
               Size: 6,581,574 bytes

Downloading using a HP-UX FTP server:

  1. Logon on to the ftp server.
  2. Create a directory:

    Note: This example uses /opt/firmware/sx2000/4_2

       # mkdir /opt/firmware/sx2000/4_2
    
  3. Copy the m_ipf_sx2000_4_2.tar.gz file to the /opt/firmware/sx2000/4_2 directory.
       # cp m_ipf_sx2000_4_2.tar.gz /opt/firmware/sx2000/4_2/
    
  4. Change directory to /opt/firmware/sx2000/4_2:
       # cd /opt/firmware/sx2000/4_2
    
  5. Execute the command below to un-compress the gzip file:
       # gunzip m_ipf_sx2000_4_2.tar.gz
    
  6. Execute the command below to un-tar the tar bundled files
       # tar -xvf m_ipf_sx2000_4_2.tar
    
  7. Run the cksum command and verify the checksum values match those below:
       # cksum *.frm
    	1522282204 12584000 hm_ipf.009.066.000.frm
    	1708795058 806112 hm_pdhc.023.003.040.frm
    	745210753 1467601 m_ap_lpm_0.002.003.000.frm
    	3226057884 1467601 m_ap_lpm_1.002.003.000.frm
    	24217201 836134 m_cell_jtag.001.002.000.frm
    	1184188901 1467263 m_cell_lpm.001.002.000.frm
    	3049656127 5246318 m_cell_pdh.001.005.001.frm
    	1111976898 827778 m_cio.001.002.000.frm
    	524948298 476568 m_ed.002.011.000.frm
    	1909090839 821051 m_emmux.001.000.000.frm
    	2832456748 1463679 m_gpm.001.002.000.frm
    	2450051573 1463679 m_io_lpm_0.001.004.000.frm
    	2235512716 1463679 m_io_lpm_1.001.004.000.frm
    	976583096 6581574 m_mp.004.002.003.frm
    

Installation:

NOTE: These installation instructions are for the MP Initiated firmware update process.

Firmware Update Procedure:
The procedure described in this section works for firmware updates regardless of which entities need to be updated except as noted below.

CAUTION: Prior to starting the firmware update, shutdown all operating systems on all partitions within the complex for the rx8640 or within the cabinet for the rx7640

  1. Make sure all update files (*.frm) are available to the MP LAN on an FTP server.
  2. Login to the MP user interface.
  3. Enter "CM" to access the Command Menu.
  4. Enter "PE" to access the Power Menu.
  5. Turn power off to the "Complex" for rx8640. Turn power off to the "Cabinet" for rx7640.
  6. Type "<ctrl> b" to return to the MP Main Menu.
  7. Enter "FW" to select the Firmware Update Utility.
  8. Select option "0 " for FTP Server.
  9. Follow the instructions to enter the FTP Server IP, Directory, User Name, and Password
  10. A list of files and the associated entities available for update will be displayed. Type "a <cr>" to select all files (This assumes there is only one file for each entity in the FTP directory).
  11. Select a target entity(s) for each file as prompted. Select "a" for 'All' entities, specify individual entities, or "s" for 'Skip' if all the entity revisions match the file revision. NOTE: Step 11 is repeated for each file in the FTP directory.
  12. Type "y "to begin the update process.
    The MP checks the state of each entity selected. An error will be returned if any entity is in a state such that the firmware update cannot be executed
    If there are no errors all the entities will be updated. This may take up to 20 minutes if all entities were selected
  13. The utility will display a prompt to reset each entity that was updated. DO NOT RESET THE ENTITIES. Answer "n" to all prompts to reset.
  14. Verify that each entity updated has a status of 'PASSED'.
  15. Reflash any entity not having a state of 'PASSED'.
  16. Once all entities updated have a state of 'PASSED', AC power cycle the Server and the SEU Cabinet (if present) as follows:
    1. Remove all AC power cords from the SEU (if present)
    2. Now remove all AC power cords from the rx8640 or rx7640
    3. Wait 30 seconds
    4. Insert the rx8640/rx7640 AC power cords
    5. Now insert the SEU AC Power cords (if present)
  17. The MP will begin the boot up process once AC Power is re-applied to the server. After the MP has booted, login to the MP user interface. Type "cm" to enter the Command Menu and then run the "sysrev" command. Using the sysrev output verify the new firmware revisions.

COMPATIBILITY/INTEROPERABILITY:
None

DETERMINING CURRENT VERSION:

  1. Login to the MP.
  2. Enter "CM" to access the Command Menu.
  3. Run "sysrev" to display system revision information.
MP:CM> sysrev

    Cabinet firmware revision report

PROGRAMMABLE HARDWARE :

System Backplane :    GPM         EMMUX
                  -----------  -----------
                  001.002.000  001.000.000

IO Backplane     :  IO_LPM-0     IO_LPM-1
                  -----------  -----------
                  001.004.000  001.004.000

Core IO          :   CIO-0        CIO-1
                  -----------  -----------
                  001.002.000  001.002.000

                    CELL_LPM    CELL_JTAG     CELL_PDH
                  -----------  -----------  -----------
Cell 0 :          001.002.000  001.002.000  001.005.001
Cell 1 :          001.002.000  001.002.000  001.005.001

FIRMWARE:

Core IO
  MP-0          :  B.004.002.003
  ED-0          :    002.011.000
  MP-1          :  B.004.002.003
  ED-1          :    002.011.000

Cell 0
  PDHC          :  B.023.003.040 - Active
  PDHC          :  B.023.003.040 - Pending
  IPF_FW        :    009.022.000
  IPF_FW        :    009.066.000 - Active
Cell 1
  PDHC          :  B.023.003.040 - Active
  PDHC          :  B.023.003.040 - Pending
  IPF_FW        :    009.022.000
  IPF_FW        :    009.066.000 - Active

KNOWN ISSUES:

  • It is not always possible to capture MCA information in NVRAM when a 2nd MCA occurs before the first MCA has been cleared from NVRAM by either a user operation or the OS clearing these error records in the boot process.

    Examples of configurations and/or conditions that have experienced this issue are:
    • A partition is booted with vPars, then transitions to nPars, then has 2 Global MCAs in nPars mode and the 2nd MCA occurs before the 1st MCA information can be transferred from NVRAM to disk.
    • Systems with configurations that utilize a lot of NVRAM space.
  • In a global MCA case under Windows, abnormal behavior may result if the VGA card for the partition is installed on IO attached to a cell that is not the lowest numbered cell in the partition. This will manifest as a hang in the OS crash dump handler following a global MCA.
  • Should this occur, reset the partition from the CM> menu to return the system to normal operation. Error logs available from EFI via the 'errdump' command and after Windows has booted are valid. However, there may be issues surrounding the crash dump information. Repeated encounters of this issue may be avoided by relocating the VGA card to IO associated with the lowest numbered cell in the partition.
  • Typing "<ctrl> s" on a local or telnet console and then rebooting the system causes the next boot to stall at various places until "<ctrl> q" is pressed on the same console session. The user has no indication as to why the system is not booting. Typing "<ctrl> q" on the original telnet session, closing the original telnet session or rebooting the MP corrects the stall. If multiple telnet sessions are present, typing a "<ctrl> s" on the active terminal halts output on the active terminal, but the other consoles can gain access to the console and perform a few commands before the stall occurs. Once stalled, recover by typing "<ctrl> q" on the same console that issued the "<ctrl> s".
  • For Montecito processors: In the event of a processor module power failure, all processor modules residing on the same processor interface bus will be powered down. Should such a power fault occur, the event logs will indicate the failed processor module with a "CPU_POWER_MODULE_FAULT" event log. If that processor interface bus is populated with two processor modules, both modules will register faults with this same event code in a manner similar as is shown below:
      PDHC 0,0   *5 a4801a8b00e00000 00ffff00ff00ff11 CPU_POWER_MODULE_FAULT
                                                      09/12/2006 09:04:13
      PDHC 0,0   *5 a4801a8b00e00000 00ffff00ff01ff11 CPU_POWER_MODULE_FAULT
                                                      09/12/2006 09:04:16
    

    In this example, only the failed module listed first via progression of time in the event logs needs to be replaced, as the second fault indicator is a direct system action that results from the real failure. The socket location of the processor module to be replaced is displayed in bits [19:16] of the event data field. The '0' above indicates that the processor module located in socket 0 is the source of the failure and should be replaced. Processor modules residing on the opposite FSB interface of the affected cell will reboot and function normally.

  • Door bell cancelling of multiple concurrent PCI OLRs in a SEU chassis may fail to cancel. When cancellation works correctly, the PCI slot LED will not flash.
  • When a user presses door bells on multiple PCI Slots to initiate OLR on multiple PCI cards, but then decides to cancel all OLRs by pressing the same door bells again, some of the cancellations may fail. The cancellation failures result in the PCI LEDs flashing, and the affected PCI cards will be suspended followed by an OLA of the same card by the OS. This issue has only been observed with the SEU chassis.

    Workaround: Do not attempt multiple, concurrent OLRs with slots in an SEU chassis. If multiple OLRs have already been queued and cancellations have also been queued, some slots may still flash the PCI LEDs. In this case, no action needs to be taken by the user. The affected slots will be suspended and then resumed by the OS.

  • Following are the known issues in MP Logging feature:
    • Only an explicit "X" command at the MP main menu will capture an MPEL logout event. Terminating the connection by terminating the telnet client will not allow a logout event to be generated.
    • Web session timeout and a subsequent user logout is not captured as an MPEL logout event.
    • A forced disconnect of users by using "DI" command or Web-Disconnect is not captured as an MPEL logout event.
    • An "MP FW Activate" event will not be logged if the MP is reset outside of the FWU Utility.
    • The "D" command for dumping all the MPEL logs will always start from the oldest log and proceed towards newest.
    • There is considerable latency in log navigation when the MPEL buffer is full and there are filters enabled for navigating through the MPEL records.
    • The following matrix shows the action events that are captured as MPEL logs -
                     | Telnet | SSH | Serial | Web |
    ------------------------------------------------
    Login            |  YES   | YES |  YES   | YES |
    Logout           |  YES   | YES |  YES   | YES |
    Login Fail       |  YES   | YES |  YES   | YES |
    MP FWU           |  YES   | YES |  YES   | NA  |
    MP FW Activate   |  YES   | YES |  YES   | NA  |
    Console Access   |  YES   | YES |  YES   | NO  |
    Log clear        |  YES   | YES |  YES   | NO  |
    
  • Following are the known issues in LDAP implementation:
    • In the MP LDAP command Security Group Administrations, option (g), if the number of partitions exceeds 5, then the Access Righs field will wrap and be displayed below the group names (This issue is present only in TUI).
    • In Extended Schema LDAP, if a user session is already open and that user is either added to another role or removed from a role, this change will not be reflected for these user sessions.
    • Any change in LDAP server parameters like IP Address and port number will not be reflected for the sessions already open. The LDAP connection is initialized only when the session is authenticated during login.
  • The creator product name still remains as "MP Web Console" as reported by the snmpwalk command
  • iLO and Reflection X don't play well together
  • When the MP hostname is set to a string of more than 60 characters, occasionally the MFW encounters a "Branch Through Zero" error.
  • The error message displayed when an invalid DNS domain name is entered is inaccurate. The actual condition for a valid MP DNS domain name is that DNS Domain Name is made up of components, each separated by a period. Each component must start with a letter, end with a letter or digit, and have only letters, digits or hyphens as interior characters. MP DNS Domain Name cannot exceed 127 characters.
  • To modify an all partition user's partition access privilege to selected partitions only, select each of the partitions one by one sequentially when
    " Enter More partitions (Y/[N]) ? :"
    message is displayed (while editing a user using so command) and explicitly set the access rights for each partition (either grant the required rights or select 'none' of the rights if access rights to a specific partition has to be removed)

DISCLAIMER: The information in this document is subject to change without notice. Hewlett-Packard makes no warranty of any kind with regard to this material, including, but not limited to, the implied warranties of merchantability and fitness for a particular purpose. Hewlett-Packard shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing, performance, or use of this material. This document contains proprietary information that is protected by copyright. All rights are reserved. No part of this document may be reproduced, photocopied, or translated to another language without the prior written consent of Hewlett-Packard Company.

(C) 1994-2010 Hewlett-Packard Development Company, L.P.

SUPERSEDES HISTORY:

ENHANCEMENTS:

Version 4.1:

  • Added support for Microsoft Windows Server 2008 Enterprise and Data Center
  • LDAP (Extended and Schemaless)
  • MP login, logout and console access logging (MPEL)
Version 4.0:
  • MP LAN supports DDNS/DHCP.
  • PCIXe support for partitions with Montecito and Montvale processors.
  • Web GUI support.
  • Integrity Lights Out Advanced KVM Card support (AD307A).
  • Integrity Lights Out web GUI with web console and remote integrated console.
  • Dynamic nPartition support for on HP-UX 11i v3 OE Update Release for Sept 2007 or later.
  • Support for 8GB DIMMs.
Version 3.1:
  • None
Version 3.0:
  • Version 3.0 supports memory migration on HP-UX 11i v3 and Dual-Core hyperthreading in vPars A.05.01.
Version 2.0:
  • Enabled support for Dual-Core Intel(R) Itanium(R) processors
  • Tape boot is supported with HP-UX 11i v2 (HWE0609)
Version 1.0
  • none

FIXES:

Version 4.1:

Core IO MP Firmware B.004.001.001 Contains the following fixes

  • MP SSH extends support to openSSH clients that use SSH Protocol 2.0 which can use more than 22 character strings during protocol version exchange.
  • In the Web GUI Administrations page clicking beside the Administration page link now works correctly
  • Web GUI now partially supports Mozilla Firefox with iLO.
Cell IPF FW 009.048.000 Contains the Following Fixes:
  • Intel PAL 9.68 for Montecito and PAL 1.14 for Montvale
  • The Montecito and Montvale PAL firmware includes fixes for a infrequent condition that may result in a fatal Machine Check instead of a correctable Machine Check.
  • If a cell is dynamically added to a partition (COLAD) and at some later time is involved in a partition-wide Machine Check, the dynamically added cell will halt after Machine Check processing and does not join the partition. A reset of the partition from the MP will enable the partition to boot properly. The fix will prevent the added cell from halting in the first place.
  • In certain vPars configurations, a vPar may not recover normally from an HP-UX crash or panic due to a pending interrupt. An nPar reboot is required to correct the hang. The fix prevents the interrupt hang and will allow the vPar to be reset with no side-effects to other vPars.
  • In some hardware configurations, repeated reboot attempts may infrequently result in a Break Instruction fault in the HP-UX bootloader. The fix prevents the cause of the Break fault.
  • PCI online add of a Smart Array P800 controller could fail infrequently. This issue has been fixed.
  • In certain rare memory system failure cases and if the cell includes deconfigured processors, the system may continuously reboot to EFI. The fix will prevent the continuous reboot loop.
  • Enhanced LANBOOT to accept command line options instead of requiring the use of dbprofile.
  • In certain customer network configurations, LANBOOT may timeout on sx2000 systems using LAN cards A7982A, A7011A. The lanboot timeout has been increased.
  • The PCIe AD221A did not appear in EFI "info io" output. Thisissue has been corrected.
  • After a PCI online add in vPars mode, the slot attention LED was not being cleared. This has been fixed.
  • Using certain Intel EFI drivers, an ftp "put" operation from EFI may fail. The fix comprehends new status from these drivers and will prevent the ftp "put" failure.
  • If a cell is deleted online and then power-cycled, a critical event is generated during what is actually normal and expected activity. The severity of the event has been lowered to match expected behavior.
  • In certain configurations, the existence of permanent correctable single-bit memory errors may cause an Alternate Data TLB fault during the boot of HP-UX. This issue has been fixed.
  • Access to the EFI boot manager menu items will be prevented in vPars mode.
  • After a non-fatal I/O error occurred during system initialization, the error was not completely cleared which caused subsequent extraneous events in the Forward Progress Log. This has been fixed.
  • If the highest numbered Montecito processor is deconfigured, the EFI "info cpu" output would longer show whether P-states are enabled or not. The "info cpu" output for Montecito processors is now complete regardless of whether any processors are deconfigured or not.
Version 4.0:

Cell IPF FW 009.022.000 Contains the Following Fixes:

  • During global MCA processing, if a link between an IO chassis and the cell is lost, the cell with missing SBA link will self-reset while other cells are still processing the global MCA. The self-reset breaks coherency domain and causes a nested, global MCA to the remaining cells. Firmware no longer resets the cell with the missing SBA link which allows the remaining cells to complete MCA processing and reboot.
  • When a PCI card is plugged into a PCI-X slot that is empty at global cold reset, but before the vPar that owns the slot is launched or reset, the vPar would MCA. This involved vPars configurations only. Firmware now correctly programs the PCI card for this corner case and no longer causes MCA.
  • A GeneralExceptionFault occurs when USB devices are connected to the Dual Function PCI Add-in card (VGA/USB) A6869A, and there is at least 1 EBC driver used in the partition. The system may MCA or fault during EBC driver initialization. The driver may also not function properly which results in boot failures. EBC drivers are used by the following cards: AB278A/B, AB379A/B, AB429A, AD300A, AE311A Firmware now provides the correct runtime environment for EBC drivers.
  • When a PCI power fault occurs in HPUX with PCI Error Handling enabled, firmware causes a nested fault and crashes HPUX. If the system is in vPars mode, the affected vPar cannot be reset. A global cold reset is required to clear the error condition. Firmware now correctly handles PCI power fault conditions which enables HPUX to handle and recover from the error.
  • When a cell with a single processor has the processor deconfigured, the cell cannot be used, even when the failed processor is replaced. To re-enable the new, but deconfigured, processor, another processor must be temporarily added to the cell. The cell with 2 processors can now boot to EFI and allows manual reconfiguration of the deconfigured processor. The temporary processor can now be removed and the single-CPU cell can boot again. Firmware now automatically re-enables the replacement processor if no other faults exists in the single-CPU cell.
  • Whenever HPUX 11.31 is in crash dump because of a panic, MCA or TOC, the crash dump CPU may call firmware to reset the hard partition while other CPUs are still gathering error logs in firmware. This can result in a deadlock condition which causes the nPar to hang and not automatically recover. Previously, to recover, the nPar had to be manually reset to clear the hang. Firmware now ensures the deadlock cannot occur.

Core IO MP Firmware B.003.001.002 Contains the Following Fixes:

  • The front panel cell LED will now flash Amber when a CPU is deconfigured on failure.
  • Previous versions of MP firmware displayed the following errors when performing a DF command option D, A (all):
          ERROR: There has been an error retrieving FRU
                 RtnCodeBusBoyDevNotPresent
    
    This error repeated once for each Montecito processor installed, as those processor's power modules (PPMs) do not have FRU data, whereas previous processors did. The new version does not attempt to retrieve that data from PPMs that do not have FRU data.
  • HPUX parstatus no longer displays incorrect fan status count after a failed fan is replaced.
  • After running the MP>dc command, telnet sessions would show up as enabled when they were actually still disabled. The MP>dc command now correctly enables telnet to the MP.
  • If a PCI latch status changes on a PCI-X slot in an IOX while a cell connected to that I/O chassis is unavailable (e.g., is pulled out or has no AC power), the previous latch state will be used on the next cell boot. This will result in one of two cases for an occupied slot, depending on the original latch status:
    1. If the latch is physically closed, an IODISC_PCI_LATCH_OPEN event will be seen. At EFI, the slot state will be shown as "HW Deconf".
    2. If the latch is physically open, an IODISC_LBA_SLOT_POWER_ERR event will be seen. At EFI, the slot state will be shown as "Failed Power".
      Both cases may inhibit boot. PCI latch status is now always updated when a cell is turned on.
  • Improvements made to MP network security for Web access.
  • Improperly disconnecting an SSH console session during high traffic will no longer lead to a hang of the partition's console.
Core IO Event Dictionary 002.008.000 Contains the Following Fixes:
  • Modifications were made to the Event Dictionary file which ensure corresponding Events that changed in firmware are accurately reported.
Version 3.1:

Cell IPF FW 007.048.000 Contains the Following Fixes:

  • In some rare circumstances, servers with Dual-Core Intel (A) Itanium (R) processors could experience a level 3 cache error, most often in the level 3 Tag, causing the partition to MCA. System firmware has been updated to correctly handle these level 3 cache errors.
    System Backplane GPM 001.002.000 Contains the Following Fixes:
  • Previously, a timing issue on the HP rx8640 Server Expansion Unit (SEU) power up could sometimes cause the GPM to select the wrong master MP in the SEU. This has now been corrected.
  • Previously, in some rare cases, a single Bulk Power Supply failure on an rx8640 could cause the cabinet to power off. This has now been corrected.
Version 3.0:

Cell IPF FW 007.044.000 Contains the Following Fixes:

  • A new feature was added to support automatic boot retry on diskless systems.
  • During boot, a link failure would generate an "ERROR: Deleted edge". This event has been changed to "WARNING: Deleted Edge." This is reported in the forward progress logs as "ARF_C2C_LINK_DOWN"
  • Due to a short power on delay during system initialization of IO, the A6826A (2p 2G Fibre Channel) card would fail and not discover attached drives. A delay was added within firmware that allows more time for IO discovery at system initialization to correct this issue.
  • When an MCA occurs on the primary VGA Obsidian card under Windows, a memory dump will not be collected. This has been fixed.
  • Error logs written to the EFI disk in vPars mode had an incorrect header file size. This file header size is now correctly allocated and the logs can be viewed successfully from the MCA Analyzer tool.
  • A change was made to EFI to support the Altiris PXE server for Windows.
  • Servers could experience poor IPF HPUX vPars performance whenever there were unassigned CPUs. Unassigned CPUs in no longer cause vPars performance issues.
Core IO MP Firmware B.002.005.010 Contains the Following Fixes:
  • Previously, the MP:CM> DF (Display FRU) Option D would incorrectly display a valid CPU power pod module's checksum as invalid. Valid checksums are no longer reported as invalid.
  • The "E" alert on the Virtual Front Panel (VFP) screen could be displayed inappropriately. The "E" alert is now cleared when the partition is reset.
  • Previously, the MP telnet connection could hang when q was pressed for main menu help on servers with an MP hostname longer than 11 characters. Hostnames longer than 11 characters no longer cause abnormal behavior.
  • Overtemp conditions would only notify cell 0's operating system to shut down. An overtemp condition now signals all cells' operating systems to shut down.
Core IO Event Dictionary 002.006.000 Contains the Following Fixes:
  • Modifications were made to the Event Dictionary file which ensure corresponding Events that changed in firmware are accurately reported.
Version 2.0:

Cell IPF FW 005.024.000 Contains the Following Fixes:

  • During extensive lab testing, HP discovered abnormal behavior that may appear under certain software code sequences on Itanium processors prior to the Intel Dual Core Itanium processor (Montecito). Firmware for Montecito-based systems does not have this exposure. The issue may occur with a small subset of Correctable Machine Check (CMC) events when PAL firmware is executed and eventually hands off to SAL firmware. SAL firmware fails to preserve pre-existing floating point registers in it's handling of the CMC event, potentially resulting in unexpected operation due to a corrupted register state. This behavior has not been reported outside the lab environment.
  • The HP F5101B USB external Floppy Drive was not recognized at the EFI shell and therefore did not function when connected to the USB ports contained on the A6869B USB/VGA PCI Card). The HP F5101B USB external Floppy Drive is now recognized at EFI.
  • The HP DVD USB external drive dvd640e (i.e. Q6764A), with a DVD present in the drive when connected to the A6869B USB/VGA PCI Card during an EFI scan caused the partition to hang indefinitely until a reset of the partition was performed. The device no longer causes a partition hang.
  • Previously, using the TOC command when there were 2 or more vPars hung in the 'Load' state could render the console unable to respond to any user input, and the system unable to automatically reset after the TOC. The TOC functionality has been corrected.
  • Re-creating a genesis profile on an rx7640 configuration with 2 Cells and a single Core IO caused HP-UX to panic during boot. Re-creating a genesis profile no longer causes HP-UX to panic.
  • Excessive memory errors at run-time could cause the Page Deallocation Table (PDT) to fill and result in an OS panic in rare situations. A full PDT no longer causes an OS panic.
  • Using HP-UX, a PCI remove/replace or add led to a dump failure if a global MCA hit prior to a reboot after the PCI operation. If a global MCA occurred after an online PCI remove/replace or add had been performed on a partition, a 2nd MCA would occur prior to the OS dump process resulting in the dump's failure. The dump now completes.
Cell PDHC Firmware B.023.003.030 Contains the Following Fixes:
  • The MP command PS option C could incorrectly display a CPU Module or Hotswap fault. CPU Module and Hotswap faults are now correctly reported.
  • After an AC cycle and before Cabinet power up (48v), the Sysrev command incorrectly displayed the active IPF_FW Image. The correct image was displayed after Cabinet power up. The IPF_FW revision is now correctly reported by sysrev prior to cabinet power up.
Core IO MP Firmware B.002.004.009 Contains the Following Fixes:
  • After removing the PCI power supply, the PS option T command showed the IO chassis populated (though not enabled and power not ok). The populated flag (*) for power status is now driven by lpm ready, rather than backplane present. That means all VRMs / PCI Power Supplies need to be present for the populated flag to be ON (*). If any VRM or PCI Power Supply is missing, the * will be off, indicating the power system is not fully populated for that IO chassis.
Core IO Event Dictionary 002.003.000 Contains the Following Fixes:
  • Additional modifications were made to the Event Dictionary file which ensure corresponding Events that changed in firmware are accurately reported.
EFI Drivers:
  • Running the "SEARCH" command at the EFI shell prior to a boot attempt through a Fibre Channel HBA would cause a boot failure and the following message:
    'Load of HP-UX Primary Boot: 1/0/1/1/0/4/0.8.0.255.0.6.0 failed: Not
    Found'
    
    (NOTE: The path displayed in the error message is the actual path which was attempted in the boot.) This affects 2Gb Fibre Channel HBA products A6826A, A9782A, A9784A, and AB465A with EFI driver version 1.47 or earlier and 4Gb Fibre Channel HBA AB378A with EFI driver version 1.01. A fix for this issue is now available in EFI driver version 1.49 which is available via the web at http://www.hp.com
Version 1.0:
  • None

FEEDBACK:

As we are continuing to improve the firmware management process we welcome your feedback on this document and on the firmware update process.
TEAM-FWupdateFeedback@external.groups.hp.com


Version:4.2(MPI Initiated) (18 Oct 2010)
Version:4.1(MP Initiated) (22 Sep 2008)
Version:4.0(MP Initiated) (21 Nov 2007)
Version:3.1(MP Initiated) (19 Nov 2007)
Version:3.0 (27 Feb 2007)
Version:2.0(MP Initiated) (20 Sep 2006)
Version:1.0(MP Initiated) (6 Apr 2006)
Type: Firmware (Entitlement Required) - System
Version: 4.2(MPI Initiated)(18 Oct 2010)
Operating System(s):
HP-UX 11.23
HP-UX 11.31 (IA)
HP-UX 11.x
Microsoft Windows Server 2003 64-Bit Edition
Microsoft Windows Server 2008 Itanium
OpenVMS v8.3
OpenVMS v8.3-1H1
Red Hat Enterprise Linux 4 (Itanium)
SUSE Linux Enterprise Server 10 (Itanium)

Description

Firmware release for HP rx8640, rx7640, and Server Expansion Unit (SEU)

Release Notes

End User License Agreements:
HPE Software License Agreement v1


HP rx8640/rx7640 and Server Expansion Unit (SEU) Firmware Release

TITLE:HP rx8640/rx7640 and Server Expansion Unit (SEU) Firmware Release

VERSION: rx8640/rx7640 and SEU (AB301A) Firmware Version 4.2

Programmable Hardware:

  System Backplane GPM   	:   001.002.000
  System Backplane EMMUX 	:   001.000.000
  IO Backplane IO_LPM-0  	:   001.004.000
  IO Backplane IO_LPM-1  	:   001.004.000
  PCIe IO Backplane IO_LPM-0	:   002.003.000
  PCIe IO Backplane IO_LPM-1	:   002.003.000
  Core IO CIO            	:   001.002.000
  Cell_LPM               	:   001.002.000
  Cell_JTAG              	:   001.002.000
  Cell_PDH               	:   001.005.001
Firmware:
  Core IO MP             	: B.004.002.003
  Core IO ED			:   002.011.000
  Core IO SCSI FW             	:   001.003.035.070
  Core IO SCSI EFI Driver	:   001.005.004.000
  Cell PDHC              	: B.023.003.040
  Cell IPF_FW            	:   009.066.000

DESCRIPTION:
This firmware release supports the HP rx8640/rx7640 servers with the following processors:

IPF Montvale CPUs:

  • 1.6GHz 24MB cache dual-core Intel (R) Itanium (R) processor 9000 series
  • 1.6GHz 18MB cache dual-core Intel (R) Itanium (R) processor 9000 series
  • 1.4GHz 12MB cache dual-core Intel (R) Itanium (R) processor 9000 series

IPF Montecito CPUs:

  • 1.6GHz 24MB cache dual-core Intel (R) Itanium (R) processor
  • 1.6GHz 18MB cache dual-core Intel (R) Itanium (R) processor
  • 1.4GHz 12MB cache dual-core Intel (R) Itanium (R) processor

IPF Mad9M CPUs:

  • 1.6 GHz 6MB L3 cache single-core Intel (R) Itanium (R) processors

UPDATE RECOMMENDATION: Recommended

HP recommends systems used in Mission Critical environments update to this revision of firmware during the next system maintenance period in order to improve system reliability.

[ ] Critical
    [ ] Panic, [ ] Hang, [ ] Abort, [ ] Corruption, [ ] Memory Leak,
    [ ] Other _______________
    [ ] Hardware Enablement, [ ] Software Enablement
[ ] Required
[X] Recommended
[ ] Optional
    [ ] Hardware Enablement, [ ] Software Enablement [ ] non-critical
[ ] Initial Customer Release

SUPERSEDES: 4.1

PRODUCT MODEL(S):

  • HP Integrity rx7640 8-socket Enterprise Server
  • HP Integrity rx8640 16-socket Enterprise Server
  • HP Integrity rx8640 16 Server Expansion Unit (SEU)

OPERATING SYSTEMS:

For dual-core Intel (R) Itanium (R) processor, Montvale:

  • HP-UX 11i v2 OE Update Release for June 2007 or later; Support for vPars A.04.04
  • HP-UX 11i v3 OE Update Release for Sept 2007 or later; Support for vPars A.05.02
  • Microsoft Windows Server 2003 64-Bit Edition
  • Microsoft Windows Server 2008 Itanium
  • Microsoft Windows Server 2008 Enterprise and Data Center
  • Microsoft Windows Server 2008 SP2 Enterprise and Data Center
  • Microsoft Windows Server 2008 R2 Enterprise and Data Center
  • OpenVMS 8.3-1H1
  • OpenVMS 8.4
  • SUSE Linux Enterprise Server 10 Service Pack 1

For dual-core Intel (R) Itanium (R) processor, Montecito:

  • HP-UX 11i v2 OE Update Release for Sept 2006 or later; Support for vPars A.04.02
  • HP-UX 11i v3; Support for vPars A.05.01
  • Microsoft Windows Server 2003 64-Bit Edition
  • Microsoft Windows Server 2008 Itanium
  • Microsoft Windows Server 2008 Enterprise and Data Center
  • Microsoft Windows Server 2008 SP2 Enterprise and Data Center
  • Microsoft Windows Server 2008 R2 Enterprise and Data Center
  • OpenVMS 8.3
  • RedHat Enterprise Linux 4 Update 4
  • RedHat Enterprise Linux 5 Update 5
  • SUSE Linux Enterprise Server 10

For single-core Intel (R) Itanium (R) processor, Mad9M:

  • HP-UX 11i v2 OE Update Release for March 2006 or later; Support for vPars A.04.01
  • HP-UX 11i v3; Support for vPars A.05.01
  • Microsoft Windows Server 2003 64-Bit Edition
  • Microsoft Windows Server 2008 Itanium
  • Microsoft Windows Server 2008 Enterprise and Data Center
  • Microsoft Windows Server 2008 SP2 Enterprise and Data Center
  • Microsoft Windows Server 2008 R2 Enterprise and Data Center

LANGUAGES:
International English

ENHANCEMENTS:

  • In webGUI, Partition names are printed along with Partition numbers.
  • MP webGUI, vKVM and vMedia features now work in a NAT environment
    NOTE: A new option for the SA command has been added -
    "U - Hostname in Web URLs"
    This has to be set to "Enabled" for the MP to work in a NAT environment
    This has to be set to "Disabled" for the MP to work in a non-NAT environment
  • Numerous improvements have been made to the MP WebGUI help pages.
  • Improvements have been made to MP user administration.

FIXES:

Cell IPF FW 009.066.000 Contains the Following Fixes:

  • Certain cell/IO combinations do not boot Windows (STOP CODE A5). All supported cell/IO combinations should now boot.
  • In rare cases, an IPF vPar panic could occur when another vPar was being rebooted. This is now fixed.
  • With Windows, cab/bay/chassis is always displayed as 0/0/0 for embedded device in IO chassis number one. The correct path is now displayed.
  • Previously, not all LUNs can be seen from EFI in some large SAN configurations. All LUNs are now visible.

Core IO MP Firmware B.004.002.003 Contains the Following Fixes:

  • MP IP addresses ending in .13 cause failures when using the iLO GUI. These addresses now work.
  • The MP telnet server now correctly handles the Telnet NOP.
  • The MP is unreachable from 192.168.0.0 subnet when netmask is other than 255.255.255.0. This now works.
  • Issues could occur when launching vMedia from 2 different browser windows on the same client machine. This has been corrected.
  • After recovery from Master MP failover state, the LDAP configuration could be lost. This has been fixed.
  • CM inactivity timeout disabled by repeated MP login attempts no longer occurs.
  • Local user web sessions were not closed on a Security Configuration reset. These are now closed during this operation.
  • Deleting a user now closes all connections previously opened by the user.
  • Users can no longer login to iLO with blank/invalid credentials when LDAP is enabled.
  • Anonymous logon is now enabled in Microsoft Active Directory, enabling LDAP-lite to work.

PREREQUISITES:
None

INSTALLATION INSTRUCTIONS:

Please review all instructions and the Hewlett-Packard support terms and conditions for precautions, scope of license, restrictions, and limitation of liability and warranties before installing this patch. It is important to read and understand these instructions completely before beginning. This can determine the success of completing the firmware update.

Unpacking the downloaded file (PC or HPUX and Checksum Values):

Firmware files may be obtained from several locations. When downloading firmware files, all files should be stored in the same location (path). This allows a firmware update of all entities during one firmware update session.

Note: Always download the files to your PC in Binary Mode.

Files may be located via a web browser by:

  1. Accessing the URL http://www.hp.com
  2. Then choose "Software & driver downloads"
  3. Enter the product number (Example: rx8640)
  4. Then choose "Cross operating system (BIOS, Firmware, Diagnostics, etc.)"
  5. Select the package from the list to begin download

The following tools will be required for firmware updates:

  • Network console connection to a MP console port.
  • Laptop and 9 pin RS-232 serial cable if network console is not available.
  • Windows XP or HP-UX FTP server with firmware file from site listed above.

Downloading using a Windows FTP Server:

Note: The Microsoft Windows environment which the file is going to be downloaded to must have WinZip installed to un-package the m_ipf_sx2000_4_2.tar.gz file.

  1. FTP m_ipf_sx2000_4_2.tar.gz to an empty folder.
  2. Open a Windows Explorer window, select the drive and folder that m_ipf_sx2000_4_2.tar.gz was downloaded to.
  3. Double click on the m_ipf_sx2000_4_2.tar.gz file to start the WinZip program and the un-packaging process.
  4. The WinZip program will ask if the file should be decompressed into an empty folder and opened. Click on the "Yes" button.
  5. The WinZip program will start and display all of the files within the m_ipf_sx2000_4_2.tar.gz file. Click on the "Extract" button.
  6. WinZip will ask for the folder that the files should be extracted to. Using the Folders/Drives portion of the window select the desired location and then click on the "Extract" button.
  7. Using the Windows Explorer window, select the drive and folder where m_ipf_sx2000_4_2.tar.gz was downloaded.
  8. Individually "Right Click" on each of the files located in this folder and select "Properties". Verify that each file has the following properties:
          hm_ipf.009.066.000.frm
               Size: 12,584,000 bytes

          hm_pdhc.023.003.040.frm
               Size: 806,112 bytes

	  m_ap_lpm_0.002.003.000.frm
	       Size: 14,67,601 bytes

	  m_ap_lpm_1.002.003.000.frm
	       Size: 14,67,601 bytes

          m_cell_jtag.001.002.000.frm
               Size: 836,134 bytes

          m_cell_lpm.001.002.000.frm
               Size: 1,467,263 bytes

          m_cell_pdh.001.005.001.frm
               Size: 5,246,318 bytes

          m_cio.001.002.000.frm
               Size: 827,778 bytes

          m_ed.002.011.000.frm
               Size: 476,568 bytes

          m_emmux.001.000.000.frm
               Size: 821,051 bytes

          m_gpm.001.002.000.frm
               Size: 1,463,679 bytes

          m_io_lpm_0.001.004.000.frm
               Size: 1,463,679 bytes

          m_io_lpm_1.001.004.000.frm
               Size: 1,463,679 bytes

	  m_mp.004.002.003.frm
               Size: 6,581,574 bytes

Downloading using a HP-UX FTP server:

  1. Logon on to the ftp server.
  2. Create a directory:

    Note: This example uses /opt/firmware/sx2000/4_2

       # mkdir /opt/firmware/sx2000/4_2
    
  3. Copy the m_ipf_sx2000_4_2.tar.gz file to the /opt/firmware/sx2000/4_2 directory.
       # cp m_ipf_sx2000_4_2.tar.gz /opt/firmware/sx2000/4_2/
    
  4. Change directory to /opt/firmware/sx2000/4_2:
       # cd /opt/firmware/sx2000/4_2
    
  5. Execute the command below to un-compress the gzip file:
       # gunzip m_ipf_sx2000_4_2.tar.gz
    
  6. Execute the command below to un-tar the tar bundled files
       # tar -xvf m_ipf_sx2000_4_2.tar
    
  7. Run the cksum command and verify the checksum values match those below:
       # cksum *.frm
    	1522282204 12584000 hm_ipf.009.066.000.frm
    	1708795058 806112 hm_pdhc.023.003.040.frm
    	745210753 1467601 m_ap_lpm_0.002.003.000.frm
    	3226057884 1467601 m_ap_lpm_1.002.003.000.frm
    	24217201 836134 m_cell_jtag.001.002.000.frm
    	1184188901 1467263 m_cell_lpm.001.002.000.frm
    	3049656127 5246318 m_cell_pdh.001.005.001.frm
    	1111976898 827778 m_cio.001.002.000.frm
    	524948298 476568 m_ed.002.011.000.frm
    	1909090839 821051 m_emmux.001.000.000.frm
    	2832456748 1463679 m_gpm.001.002.000.frm
    	2450051573 1463679 m_io_lpm_0.001.004.000.frm
    	2235512716 1463679 m_io_lpm_1.001.004.000.frm
    	976583096 6581574 m_mp.004.002.003.frm
    

Installation:

NOTE: These installation instructions are for the MP Initiated firmware update process.

Firmware Update Procedure:
The procedure described in this section works for firmware updates regardless of which entities need to be updated except as noted below.

CAUTION: Prior to starting the firmware update, shutdown all operating systems on all partitions within the complex for the rx8640 or within the cabinet for the rx7640

  1. Make sure all update files (*.frm) are available to the MP LAN on an FTP server.
  2. Login to the MP user interface.
  3. Enter "CM" to access the Command Menu.
  4. Enter "PE" to access the Power Menu.
  5. Turn power off to the "Complex" for rx8640. Turn power off to the "Cabinet" for rx7640.
  6. Type "<ctrl> b" to return to the MP Main Menu.
  7. Enter "FW" to select the Firmware Update Utility.
  8. Select option "0 " for FTP Server.
  9. Follow the instructions to enter the FTP Server IP, Directory, User Name, and Password
  10. A list of files and the associated entities available for update will be displayed. Type "a <cr>" to select all files (This assumes there is only one file for each entity in the FTP directory).
  11. Select a target entity(s) for each file as prompted. Select "a" for 'All' entities, specify individual entities, or "s" for 'Skip' if all the entity revisions match the file revision. NOTE: Step 11 is repeated for each file in the FTP directory.
  12. Type "y "to begin the update process.
    The MP checks the state of each entity selected. An error will be returned if any entity is in a state such that the firmware update cannot be executed
    If there are no errors all the entities will be updated. This may take up to 20 minutes if all entities were selected
  13. The utility will display a prompt to reset each entity that was updated. DO NOT RESET THE ENTITIES. Answer "n" to all prompts to reset.
  14. Verify that each entity updated has a status of 'PASSED'.
  15. Reflash any entity not having a state of 'PASSED'.
  16. Once all entities updated have a state of 'PASSED', AC power cycle the Server and the SEU Cabinet (if present) as follows:
    1. Remove all AC power cords from the SEU (if present)
    2. Now remove all AC power cords from the rx8640 or rx7640
    3. Wait 30 seconds
    4. Insert the rx8640/rx7640 AC power cords
    5. Now insert the SEU AC Power cords (if present)
  17. The MP will begin the boot up process once AC Power is re-applied to the server. After the MP has booted, login to the MP user interface. Type "cm" to enter the Command Menu and then run the "sysrev" command. Using the sysrev output verify the new firmware revisions.

COMPATIBILITY/INTEROPERABILITY:
None

DETERMINING CURRENT VERSION:

  1. Login to the MP.
  2. Enter "CM" to access the Command Menu.
  3. Run "sysrev" to display system revision information.
MP:CM> sysrev

    Cabinet firmware revision report

PROGRAMMABLE HARDWARE :

System Backplane :    GPM         EMMUX
                  -----------  -----------
                  001.002.000  001.000.000

IO Backplane     :  IO_LPM-0     IO_LPM-1
                  -----------  -----------
                  001.004.000  001.004.000

Core IO          :   CIO-0        CIO-1
                  -----------  -----------
                  001.002.000  001.002.000

                    CELL_LPM    CELL_JTAG     CELL_PDH
                  -----------  -----------  -----------
Cell 0 :          001.002.000  001.002.000  001.005.001
Cell 1 :          001.002.000  001.002.000  001.005.001

FIRMWARE:

Core IO
  MP-0          :  B.004.002.003
  ED-0          :    002.011.000
  MP-1          :  B.004.002.003
  ED-1          :    002.011.000

Cell 0
  PDHC          :  B.023.003.040 - Active
  PDHC          :  B.023.003.040 - Pending
  IPF_FW        :    009.022.000
  IPF_FW        :    009.066.000 - Active
Cell 1
  PDHC          :  B.023.003.040 - Active
  PDHC          :  B.023.003.040 - Pending
  IPF_FW        :    009.022.000
  IPF_FW        :    009.066.000 - Active

KNOWN ISSUES:

  • It is not always possible to capture MCA information in NVRAM when a 2nd MCA occurs before the first MCA has been cleared from NVRAM by either a user operation or the OS clearing these error records in the boot process.

    Examples of configurations and/or conditions that have experienced this issue are:
    • A partition is booted with vPars, then transitions to nPars, then has 2 Global MCAs in nPars mode and the 2nd MCA occurs before the 1st MCA information can be transferred from NVRAM to disk.
    • Systems with configurations that utilize a lot of NVRAM space.
  • In a global MCA case under Windows, abnormal behavior may result if the VGA card for the partition is installed on IO attached to a cell that is not the lowest numbered cell in the partition. This will manifest as a hang in the OS crash dump handler following a global MCA.
  • Should this occur, reset the partition from the CM> menu to return the system to normal operation. Error logs available from EFI via the 'errdump' command and after Windows has booted are valid. However, there may be issues surrounding the crash dump information. Repeated encounters of this issue may be avoided by relocating the VGA card to IO associated with the lowest numbered cell in the partition.
  • Typing "<ctrl> s" on a local or telnet console and then rebooting the system causes the next boot to stall at various places until "<ctrl> q" is pressed on the same console session. The user has no indication as to why the system is not booting. Typing "<ctrl> q" on the original telnet session, closing the original telnet session or rebooting the MP corrects the stall. If multiple telnet sessions are present, typing a "<ctrl> s" on the active terminal halts output on the active terminal, but the other consoles can gain access to the console and perform a few commands before the stall occurs. Once stalled, recover by typing "<ctrl> q" on the same console that issued the "<ctrl> s".
  • For Montecito processors: In the event of a processor module power failure, all processor modules residing on the same processor interface bus will be powered down. Should such a power fault occur, the event logs will indicate the failed processor module with a "CPU_POWER_MODULE_FAULT" event log. If that processor interface bus is populated with two processor modules, both modules will register faults with this same event code in a manner similar as is shown below:
      PDHC 0,0   *5 a4801a8b00e00000 00ffff00ff00ff11 CPU_POWER_MODULE_FAULT
                                                      09/12/2006 09:04:13
      PDHC 0,0   *5 a4801a8b00e00000 00ffff00ff01ff11 CPU_POWER_MODULE_FAULT
                                                      09/12/2006 09:04:16
    

    In this example, only the failed module listed first via progression of time in the event logs needs to be replaced, as the second fault indicator is a direct system action that results from the real failure. The socket location of the processor module to be replaced is displayed in bits [19:16] of the event data field. The '0' above indicates that the processor module located in socket 0 is the source of the failure and should be replaced. Processor modules residing on the opposite FSB interface of the affected cell will reboot and function normally.

  • Door bell cancelling of multiple concurrent PCI OLRs in a SEU chassis may fail to cancel. When cancellation works correctly, the PCI slot LED will not flash.
  • When a user presses door bells on multiple PCI Slots to initiate OLR on multiple PCI cards, but then decides to cancel all OLRs by pressing the same door bells again, some of the cancellations may fail. The cancellation failures result in the PCI LEDs flashing, and the affected PCI cards will be suspended followed by an OLA of the same card by the OS. This issue has only been observed with the SEU chassis.

    Workaround: Do not attempt multiple, concurrent OLRs with slots in an SEU chassis. If multiple OLRs have already been queued and cancellations have also been queued, some slots may still flash the PCI LEDs. In this case, no action needs to be taken by the user. The affected slots will be suspended and then resumed by the OS.

  • Following are the known issues in MP Logging feature:
    • Only an explicit "X" command at the MP main menu will capture an MPEL logout event. Terminating the connection by terminating the telnet client will not allow a logout event to be generated.
    • Web session timeout and a subsequent user logout is not captured as an MPEL logout event.
    • A forced disconnect of users by using "DI" command or Web-Disconnect is not captured as an MPEL logout event.
    • An "MP FW Activate" event will not be logged if the MP is reset outside of the FWU Utility.
    • The "D" command for dumping all the MPEL logs will always start from the oldest log and proceed towards newest.
    • There is considerable latency in log navigation when the MPEL buffer is full and there are filters enabled for navigating through the MPEL records.
    • The following matrix shows the action events that are captured as MPEL logs -
                     | Telnet | SSH | Serial | Web |
    ------------------------------------------------
    Login            |  YES   | YES |  YES   | YES |
    Logout           |  YES   | YES |  YES   | YES |
    Login Fail       |  YES   | YES |  YES   | YES |
    MP FWU           |  YES   | YES |  YES   | NA  |
    MP FW Activate   |  YES   | YES |  YES   | NA  |
    Console Access   |  YES   | YES |  YES   | NO  |
    Log clear        |  YES   | YES |  YES   | NO  |
    
  • Following are the known issues in LDAP implementation:
    • In the MP LDAP command Security Group Administrations, option (g), if the number of partitions exceeds 5, then the Access Righs field will wrap and be displayed below the group names (This issue is present only in TUI).
    • In Extended Schema LDAP, if a user session is already open and that user is either added to another role or removed from a role, this change will not be reflected for these user sessions.
    • Any change in LDAP server parameters like IP Address and port number will not be reflected for the sessions already open. The LDAP connection is initialized only when the session is authenticated during login.
  • The creator product name still remains as "MP Web Console" as reported by the snmpwalk command
  • iLO and Reflection X don't play well together
  • When the MP hostname is set to a string of more than 60 characters, occasionally the MFW encounters a "Branch Through Zero" error.
  • The error message displayed when an invalid DNS domain name is entered is inaccurate. The actual condition for a valid MP DNS domain name is that DNS Domain Name is made up of components, each separated by a period. Each component must start with a letter, end with a letter or digit, and have only letters, digits or hyphens as interior characters. MP DNS Domain Name cannot exceed 127 characters.
  • To modify an all partition user's partition access privilege to selected partitions only, select each of the partitions one by one sequentially when
    " Enter More partitions (Y/[N]) ? :"
    message is displayed (while editing a user using so command) and explicitly set the access rights for each partition (either grant the required rights or select 'none' of the rights if access rights to a specific partition has to be removed)

DISCLAIMER: The information in this document is subject to change without notice. Hewlett-Packard makes no warranty of any kind with regard to this material, including, but not limited to, the implied warranties of merchantability and fitness for a particular purpose. Hewlett-Packard shall not be liable for errors contained herein or for incidental or consequential damages in connection with the furnishing, performance, or use of this material. This document contains proprietary information that is protected by copyright. All rights are reserved. No part of this document may be reproduced, photocopied, or translated to another language without the prior written consent of Hewlett-Packard Company.

(C) 1994-2010 Hewlett-Packard Development Company, L.P.

SUPERSEDES HISTORY:

ENHANCEMENTS:

Version 4.1:

  • Added support for Microsoft Windows Server 2008 Enterprise and Data Center
  • LDAP (Extended and Schemaless)
  • MP login, logout and console access logging (MPEL)
Version 4.0:
  • MP LAN supports DDNS/DHCP.
  • PCIXe support for partitions with Montecito and Montvale processors.
  • Web GUI support.
  • Integrity Lights Out Advanced KVM Card support (AD307A).
  • Integrity Lights Out web GUI with web console and remote integrated console.
  • Dynamic nPartition support for on HP-UX 11i v3 OE Update Release for Sept 2007 or later.
  • Support for 8GB DIMMs.
Version 3.1:
  • None
Version 3.0:
  • Version 3.0 supports memory migration on HP-UX 11i v3 and Dual-Core hyperthreading in vPars A.05.01.
Version 2.0:
  • Enabled support for Dual-Core Intel(R) Itanium(R) processors
  • Tape boot is supported with HP-UX 11i v2 (HWE0609)
Version 1.0
  • none

FIXES:

Version 4.1:

Core IO MP Firmware B.004.001.001 Contains the following fixes

  • MP SSH extends support to openSSH clients that use SSH Protocol 2.0 which can use more than 22 character strings during protocol version exchange.
  • In the Web GUI Administrations page clicking beside the Administration page link now works correctly
  • Web GUI now partially supports Mozilla Firefox with iLO.
Cell IPF FW 009.048.000 Contains the Following Fixes:
  • Intel PAL 9.68 for Montecito and PAL 1.14 for Montvale
  • The Montecito and Montvale PAL firmware includes fixes for a infrequent condition that may result in a fatal Machine Check instead of a correctable Machine Check.
  • If a cell is dynamically added to a partition (COLAD) and at some later time is involved in a partition-wide Machine Check, the dynamically added cell will halt after Machine Check processing and does not join the partition. A reset of the partition from the MP will enable the partition to boot properly. The fix will prevent the added cell from halting in the first place.
  • In certain vPars configurations, a vPar may not recover normally from an HP-UX crash or panic due to a pending interrupt. An nPar reboot is required to correct the hang. The fix prevents the interrupt hang and will allow the vPar to be reset with no side-effects to other vPars.
  • In some hardware configurations, repeated reboot attempts may infrequently result in a Break Instruction fault in the HP-UX bootloader. The fix prevents the cause of the Break fault.
  • PCI online add of a Smart Array P800 controller could fail infrequently. This issue has been fixed.
  • In certain rare memory system failure cases and if the cell includes deconfigured processors, the system may continuously reboot to EFI. The fix will prevent the continuous reboot loop.
  • Enhanced LANBOOT to accept command line options instead of requiring the use of dbprofile.
  • In certain customer network configurations, LANBOOT may timeout on sx2000 systems using LAN cards A7982A, A7011A. The lanboot timeout has been increased.
  • The PCIe AD221A did not appear in EFI "info io" output. Thisissue has been corrected.
  • After a PCI online add in vPars mode, the slot attention LED was not being cleared. This has been fixed.
  • Using certain Intel EFI drivers, an ftp "put" operation from EFI may fail. The fix comprehends new status from these drivers and will prevent the ftp "put" failure.
  • If a cell is deleted online and then power-cycled, a critical event is generated during what is actually normal and expected activity. The severity of the event has been lowered to match expected behavior.
  • In certain configurations, the existence of permanent correctable single-bit memory errors may cause an Alternate Data TLB fault during the boot of HP-UX. This issue has been fixed.
  • Access to the EFI boot manager menu items will be prevented in vPars mode.
  • After a non-fatal I/O error occurred during system initialization, the error was not completely cleared which caused subsequent extraneous events in the Forward Progress Log. This has been fixed.
  • If the highest numbered Montecito processor is deconfigured, the EFI "info cpu" output would longer show whether P-states are enabled or not. The "info cpu" output for Montecito processors is now complete regardless of whether any processors are deconfigured or not.
Version 4.0:

Cell IPF FW 009.022.000 Contains the Following Fixes:

  • During global MCA processing, if a link between an IO chassis and the cell is lost, the cell with missing SBA link will self-reset while other cells are still processing the global MCA. The self-reset breaks coherency domain and causes a nested, global MCA to the remaining cells. Firmware no longer resets the cell with the missing SBA link which allows the remaining cells to complete MCA processing and reboot.
  • When a PCI card is plugged into a PCI-X slot that is empty at global cold reset, but before the vPar that owns the slot is launched or reset, the vPar would MCA. This involved vPars configurations only. Firmware now correctly programs the PCI card for this corner case and no longer causes MCA.
  • A GeneralExceptionFault occurs when USB devices are connected to the Dual Function PCI Add-in card (VGA/USB) A6869A, and there is at least 1 EBC driver used in the partition. The system may MCA or fault during EBC driver initialization. The driver may also not function properly which results in boot failures. EBC drivers are used by the following cards: AB278A/B, AB379A/B, AB429A, AD300A, AE311A Firmware now provides the correct runtime environment for EBC drivers.
  • When a PCI power fault occurs in HPUX with PCI Error Handling enabled, firmware causes a nested fault and crashes HPUX. If the system is in vPars mode, the affected vPar cannot be reset. A global cold reset is required to clear the error condition. Firmware now correctly handles PCI power fault conditions which enables HPUX to handle and recover from the error.
  • When a cell with a single processor has the processor deconfigured, the cell cannot be used, even when the failed processor is replaced. To re-enable the new, but deconfigured, processor, another processor must be temporarily added to the cell. The cell with 2 processors can now boot to EFI and allows manual reconfiguration of the deconfigured processor. The temporary processor can now be removed and the single-CPU cell can boot again. Firmware now automatically re-enables the replacement processor if no other faults exists in the single-CPU cell.
  • Whenever HPUX 11.31 is in crash dump because of a panic, MCA or TOC, the crash dump CPU may call firmware to reset the hard partition while other CPUs are still gathering error logs in firmware. This can result in a deadlock condition which causes the nPar to hang and not automatically recover. Previously, to recover, the nPar had to be manually reset to clear the hang. Firmware now ensures the deadlock cannot occur.

Core IO MP Firmware B.003.001.002 Contains the Following Fixes:

  • The front panel cell LED will now flash Amber when a CPU is deconfigured on failure.
  • Previous versions of MP firmware displayed the following errors when performing a DF command option D, A (all):
          ERROR: There has been an error retrieving FRU
                 RtnCodeBusBoyDevNotPresent
    
    This error repeated once for each Montecito processor installed, as those processor's power modules (PPMs) do not have FRU data, whereas previous processors did. The new version does not attempt to retrieve that data from PPMs that do not have FRU data.
  • HPUX parstatus no longer displays incorrect fan status count after a failed fan is replaced.
  • After running the MP>dc command, telnet sessions would show up as enabled when they were actually still disabled. The MP>dc command now correctly enables telnet to the MP.
  • If a PCI latch status changes on a PCI-X slot in an IOX while a cell connected to that I/O chassis is unavailable (e.g., is pulled out or has no AC power), the previous latch state will be used on the next cell boot. This will result in one of two cases for an occupied slot, depending on the original latch status:
    1. If the latch is physically closed, an IODISC_PCI_LATCH_OPEN event will be seen. At EFI, the slot state will be shown as "HW Deconf".
    2. If the latch is physically open, an IODISC_LBA_SLOT_POWER_ERR event will be seen. At EFI, the slot state will be shown as "Failed Power".
      Both cases may inhibit boot. PCI latch status is now always updated when a cell is turned on.
  • Improvements made to MP network security for Web access.
  • Improperly disconnecting an SSH console session during high traffic will no longer lead to a hang of the partition's console.
Core IO Event Dictionary 002.008.000 Contains the Following Fixes:
  • Modifications were made to the Event Dictionary file which ensure corresponding Events that changed in firmware are accurately reported.
Version 3.1:

Cell IPF FW 007.048.000 Contains the Following Fixes:

  • In some rare circumstances, servers with Dual-Core Intel (A) Itanium (R) processors could experience a level 3 cache error, most often in the level 3 Tag, causing the partition to MCA. System firmware has been updated to correctly handle these level 3 cache errors.
    System Backplane GPM 001.002.000 Contains the Following Fixes:
  • Previously, a timing issue on the HP rx8640 Server Expansion Unit (SEU) power up could sometimes cause the GPM to select the wrong master MP in the SEU. This has now been corrected.
  • Previously, in some rare cases, a single Bulk Power Supply failure on an rx8640 could cause the cabinet to power off. This has now been corrected.
Version 3.0:

Cell IPF FW 007.044.000 Contains the Following Fixes:

  • A new feature was added to support automatic boot retry on diskless systems.
  • During boot, a link failure would generate an "ERROR: Deleted edge". This event has been changed to "WARNING: Deleted Edge." This is reported in the forward progress logs as "ARF_C2C_LINK_DOWN"
  • Due to a short power on delay during system initialization of IO, the A6826A (2p 2G Fibre Channel) card would fail and not discover attached drives. A delay was added within firmware that allows more time for IO discovery at system initialization to correct this issue.
  • When an MCA occurs on the primary VGA Obsidian card under Windows, a memory dump will not be collected. This has been fixed.
  • Error logs written to the EFI disk in vPars mode had an incorrect header file size. This file header size is now correctly allocated and the logs can be viewed successfully from the MCA Analyzer tool.
  • A change was made to EFI to support the Altiris PXE server for Windows.
  • Servers could experience poor IPF HPUX vPars performance whenever there were unassigned CPUs. Unassigned CPUs in no longer cause vPars performance issues.
Core IO MP Firmware B.002.005.010 Contains the Following Fixes:
  • Previously, the MP:CM> DF (Display FRU) Option D would incorrectly display a valid CPU power pod module's checksum as invalid. Valid checksums are no longer reported as invalid.
  • The "E" alert on the Virtual Front Panel (VFP) screen could be displayed inappropriately. The "E" alert is now cleared when the partition is reset.
  • Previously, the MP telnet connection could hang when q was pressed for main menu help on servers with an MP hostname longer than 11 characters. Hostnames longer than 11 characters no longer cause abnormal behavior.
  • Overtemp conditions would only notify cell 0's operating system to shut down. An overtemp condition now signals all cells' operating systems to shut down.
Core IO Event Dictionary 002.006.000 Contains the Following Fixes:
  • Modifications were made to the Event Dictionary file which ensure corresponding Events that changed in firmware are accurately reported.
Version 2.0:

Cell IPF FW 005.024.000 Contains the Following Fixes:

  • During extensive lab testing, HP discovered abnormal behavior that may appear under certain software code sequences on Itanium processors prior to the Intel Dual Core Itanium processor (Montecito). Firmware for Montecito-based systems does not have this exposure. The issue may occur with a small subset of Correctable Machine Check (CMC) events when PAL firmware is executed and eventually hands off to SAL firmware. SAL firmware fails to preserve pre-existing floating point registers in it's handling of the CMC event, potentially resulting in unexpected operation due to a corrupted register state. This behavior has not been reported outside the lab environment.
  • The HP F5101B USB external Floppy Drive was not recognized at the EFI shell and therefore did not function when connected to the USB ports contained on the A6869B USB/VGA PCI Card). The HP F5101B USB external Floppy Drive is now recognized at EFI.
  • The HP DVD USB external drive dvd640e (i.e. Q6764A), with a DVD present in the drive when connected to the A6869B USB/VGA PCI Card during an EFI scan caused the partition to hang indefinitely until a reset of the partition was performed. The device no longer causes a partition hang.
  • Previously, using the TOC command when there were 2 or more vPars hung in the 'Load' state could render the console unable to respond to any user input, and the system unable to automatically reset after the TOC. The TOC functionality has been corrected.
  • Re-creating a genesis profile on an rx7640 configuration with 2 Cells and a single Core IO caused HP-UX to panic during boot. Re-creating a genesis profile no longer causes HP-UX to panic.
  • Excessive memory errors at run-time could cause the Page Deallocation Table (PDT) to fill and result in an OS panic in rare situations. A full PDT no longer causes an OS panic.
  • Using HP-UX, a PCI remove/replace or add led to a dump failure if a global MCA hit prior to a reboot after the PCI operation. If a global MCA occurred after an online PCI remove/replace or add had been performed on a partition, a 2nd MCA would occur prior to the OS dump process resulting in the dump's failure. The dump now completes.
Cell PDHC Firmware B.023.003.030 Contains the Following Fixes:
  • The MP command PS option C could incorrectly display a CPU Module or Hotswap fault. CPU Module and Hotswap faults are now correctly reported.
  • After an AC cycle and before Cabinet power up (48v), the Sysrev command incorrectly displayed the active IPF_FW Image. The correct image was displayed after Cabinet power up. The IPF_FW revision is now correctly reported by sysrev prior to cabinet power up.
Core IO MP Firmware B.002.004.009 Contains the Following Fixes:
  • After removing the PCI power supply, the PS option T command showed the IO chassis populated (though not enabled and power not ok). The populated flag (*) for power status is now driven by lpm ready, rather than backplane present. That means all VRMs / PCI Power Supplies need to be present for the populated flag to be ON (*). If any VRM or PCI Power Supply is missing, the * will be off, indicating the power system is not fully populated for that IO chassis.
Core IO Event Dictionary 002.003.000 Contains the Following Fixes:
  • Additional modifications were made to the Event Dictionary file which ensure corresponding Events that changed in firmware are accurately reported.
EFI Drivers:
  • Running the "SEARCH" command at the EFI shell prior to a boot attempt through a Fibre Channel HBA would cause a boot failure and the following message:
    'Load of HP-UX Primary Boot: 1/0/1/1/0/4/0.8.0.255.0.6.0 failed: Not
    Found'
    
    (NOTE: The path displayed in the error message is the actual path which was attempted in the boot.) This affects 2Gb Fibre Channel HBA products A6826A, A9782A, A9784A, and AB465A with EFI driver version 1.47 or earlier and 4Gb Fibre Channel HBA AB378A with EFI driver version 1.01. A fix for this issue is now available in EFI driver version 1.49 which is available via the web at http://www.hp.com
Version 1.0:
  • None

FEEDBACK:

As we are continuing to improve the firmware management process we welcome your feedback on this document and on the firmware update process.
TEAM-FWupdateFeedback@external.groups.hp.com


Revision History

Version:4.2(MPI Initiated) (18 Oct 2010)
Version:4.1(MP Initiated) (22 Sep 2008)
Version:4.0(MP Initiated) (21 Nov 2007)
Version:3.1(MP Initiated) (19 Nov 2007)
Version:3.0 (27 Feb 2007)
Version:2.0(MP Initiated) (20 Sep 2006)
Version:1.0(MP Initiated) (6 Apr 2006)

Legal Disclaimer: Products sold prior to the November 1, 2015 separation of Hewlett-Packard Company into Hewlett Packard Enterprise Company and HP Inc. may have older product names and model numbers that differ from current models.