SPI flash image by Flash Update

[hpcdkhab]

Dear All,

if I try to update the flash image:

sudo /opt/intel/mic/bin/micflash -Update /opt/intel/mic/flash/EXT_HP2_B0_0375-05.rom.smc

I get the following error message:

VERSION: Copyright 2011-2012 Intel Corporation All Rights Reserved.
VERSION: 4982-15

Intel(R) Xeon Phi(TM) Coprocessor - 0
  Flash update  : Failed; Reason: SPI flash image on Intel(R) Xeon Phi(TM) Coprocessor is corrupted
  Device status : HW ready
How can I handle the error ?

regards

Dmitry

[Frances_R_Intel]

I'm not sure you can. I will check with some folks - but is there any more you can tell me about how this occurred? Is this the first time you have tried updating flash? And can you send the output from the the micinfo command?

[hpcdkhab]

It was the secon update of the device.

Meanwhile, it have update the flush:

I have uninstalled all packages and install it one more time. The difference is that the first installation was with OFED-1.5.4.1 support and  there was a conflict with dapl-utils &compat-dapl-utils  (/usr/share/man/man5/dat.conf.5.gz). I could "avoid" this problem by installing one of the packages separatly. The Workstation does not have any Infiniband cards, so I could not start openibd service. I'm not sure it has to do with flash update issue.

The question is only whether the MIC supports the mpi without OFED ? I remember me, that I had to install the "OFED " in order to use mpi (at least  it was necessary with Alpha version) ?

But now I have following problem:

sudo /opt/intel/mic/bin/micinfo
MicInfo Utility Log

Created Tue Mar  5 23:02:59 2013


    System Info
        Host OS                  : Linux
        OS Version               : 2.6.32-279.19.1.el6.x86_64
        Driver Version           : NotAvailable
        MPSS Version             : 2.1.4982-15
        Host Physical Memory     : 24543 MB
        CPU Family               : GenuineIntel  Family 6  Model 44  Stepping 2
        CPU Speed                : 1596.000
        Threads per Core         : 2


*************************** The information below is not complete **************************
****** Please start the MPSS service and run MicInfo again to view the entire output *******


Device No: 0,  Device Name: Intel(R) Xeon Phi(TM) Coprocessor

    Version
        Flash Version            : NotAvailable
        UOS Version              : NotAvailable
        Device Serial Number     : NotAvailable

    Board
        Vendor ID                : 8086
        Device ID                : 225c
        SubSystem ID             : 2500
        Coprocessor Stepping ID  : f
        PCIe Width               : x16
        PCIe Speed               : 5 GT/s
        PCIe Max payload size    : 256 bytes
        PCIe Max read req size   : 512 bytes
        Coprocessor Model        : 0x0f
        Coprocessor Model Ext    : 0x0f
        Coprocessor Type         : 0x03
        Coprocessor Family       : 0x0f
        Coprocessor Family Ext   : 0xff
        Coprocessor Stepping     : Undefined
        Board SKU                : NotAvailable
        ECC Mode                 : NotAvailable
        SMC HW Revision          : NotAvailable

    Core
        Voltage                  : NotAvailable
        Frequency                : NotAvailable

    Thermal
        Fan Speed Control        : NotAvailable
        SMC Firmware Version     : NotAvailable
        FSC Strap                : NotAvailable
        Fan RPM                  : NotAvailable
        Fan PWM                  : NotAvailable
        Die Temp                 : NotAvailable

    GDDR
        GDDR Vendor              : NotAvailable
        GDDR Version             : NotAvailable
        GDDR Density             : NotAvailable
        GDDR Size                : NotAvailable
        GDDR Technology          : NotAvailable
        GDDR Speed               : NotAvailable
        GDDR Frequency           : NotAvailable
        GDDR Voltage             : NotAvailable

sudo service mpss status
mpss is running/opt/intel/mic/bin/miccheck

miccheck 2.1.4982-15, created 08:30:46 Dec 17 2012
Copyright 2011-2012 Intel Corporation  All rights reserved

Test 1 Ensure installation matches manifest : OK
Test 2 Ensure host driver is loaded         : OK
Test 3 Ensure driver matches manifest       : OK
Test 4 Detect all listed devices            : OK
MIC 0 Test  1 Find the device                       : OK
MIC 0 Test  2 Read device configuration file        : OK
MIC 0 Test  3 Ensure IP address is unique           : OK
MIC 0 Test  4 Ensure MAC address is unique          : OK
MIC 0 Test  5 Check the POST code via PCI           : FAILED
MIC 0 Test  5> get POST code: Error 0xc4: Service is not supported

MIC 0 Test  6 Ping the device                       : SKIPPED
MIC 0 Test  6> Prerequisite 'Ensure the device is online' failed:
MIC 0 Test  6>  The device is not online
MIC 0 Test  7 Connect to the device                 : SKIPPED
MIC 0 Test  7> Prerequisite 'Ensure the device is online' failed:
MIC 0 Test  7>  The device is not online
MIC 0 Test  8 Check for normal mode                 : SKIPPED
MIC 0 Test  8> Prerequisite 'Ensure the device is online' failed:
MIC 0 Test  8>  The device is not online
MIC 0 Test  9 Check the POST code via SCIF          : SKIPPED
MIC 0 Test  9> Prerequisite 'Ensure the device is online' failed:
MIC 0 Test  9>  The device is not online
MIC 0 Test 10 Send data to the device               : SKIPPED
MIC 0 Test 10> Prerequisite 'Check for normal mode' failed:
MIC 0 Test 10>  The device is not in normal mode
MIC 0 Test 11 Compare the PCI configuration         : OK
MIC 0 Test 12 Ensure Flash version matches manifest : SKIPPED
MIC 0 Test 12> Prerequisite 'Check for normal mode' failed:
MIC 0 Test 12>  The device is not in normal mode
MIC 0 Test 13 Ping the host                         : SKIPPED
MIC 0 Test 13> Prerequisite 'Check for normal mode' failed:
MIC 0 Test 13>  The device is not in normal mode
Status: Error 0xc4: Service is not supported

Do you have any idea, where the promble can be ?

[hpcdkhab]

It was the second update. The first one was successful.

But now I have new information for you:

I have uninstall all package and install it one more time.

It has helped. I have updated the flash. But something is wrong with the service mpss.

The output of micinfo is:

sudo /opt/intel/mic/bin/micinfo
MicInfo Utility Log

Created Tue Mar  5 23:02:59 2013


    System Info
        Host OS                  : Linux
        OS Version               : 2.6.32-279.19.1.el6.x86_64
        Driver Version           : NotAvailable
        MPSS Version             : 2.1.4982-15
        Host Physical Memory     : 24543 MB
        CPU Family               : GenuineIntel  Family 6  Model 44  Stepping 2
        CPU Speed                : 1596.000
        Threads per Core         : 2


*************************** The information below is not complete **************************
****** Please start the MPSS service and run MicInfo again to view the entire output *******


Device No: 0,  Device Name: Intel(R) Xeon Phi(TM) Coprocessor

    Version
        Flash Version            : NotAvailable
        UOS Version              : NotAvailable
        Device Serial Number     : NotAvailable

    Board
        Vendor ID                : 8086
        Device ID                : 225c
        SubSystem ID             : 2500
        Coprocessor Stepping ID  : f
        PCIe Width               : x16
        PCIe Speed               : 5 GT/s
        PCIe Max payload size    : 256 bytes
        PCIe Max read req size   : 512 bytes
        Coprocessor Model        : 0x0f
        Coprocessor Model Ext    : 0x0f
        Coprocessor Type         : 0x03
        Coprocessor Family       : 0x0f
        Coprocessor Family Ext   : 0xff
        Coprocessor Stepping     : Undefined
        Board SKU                : NotAvailable
        ECC Mode                 : NotAvailable
        SMC HW Revision          : NotAvailable

    Core
        Voltage                  : NotAvailable
        Frequency                : NotAvailable

    Thermal
        Fan Speed Control        : NotAvailable
        SMC Firmware Version     : NotAvailable
        FSC Strap                : NotAvailable
        Fan RPM                  : NotAvailable
        Fan PWM                  : NotAvailable
        Die Temp                 : NotAvailable

    GDDR
        GDDR Vendor              : NotAvailable
        GDDR Version             : NotAvailable
        GDDR Density             : NotAvailable
        GDDR Size                : NotAvailable
        GDDR Technology          : NotAvailable
        GDDR Speed               : NotAvailable
        GDDR Frequency           : NotAvailable
        GDDR Voltage             : NotAvailable

The output of miccheck is:

sudo /opt/intel/mic/bin/miccheck

miccheck 2.1.4982-15, created 08:30:46 Dec 17 2012
Copyright 2011-2012 Intel Corporation  All rights reserved

Test 1 Ensure installation matches manifest : OK
Test 2 Ensure host driver is loaded         : OK
Test 3 Ensure driver matches manifest       : OK
Test 4 Detect all listed devices            : OK
MIC 0 Test  1 Find the device                       : OK
MIC 0 Test  2 Read device configuration file        : OK
MIC 0 Test  3 Ensure IP address is unique           : OK
MIC 0 Test  4 Ensure MAC address is unique          : OK
MIC 0 Test  5 Check the POST code via PCI           : FAILED
MIC 0 Test  5> get POST code: Error 0xc4: Service is not supported

MIC 0 Test  6 Ping the device                       : SKIPPED
MIC 0 Test  6> Prerequisite 'Ensure the device is online' failed:
MIC 0 Test  6>  The device is not online
MIC 0 Test  7 Connect to the device                 : SKIPPED
MIC 0 Test  7> Prerequisite 'Ensure the device is online' failed:
MIC 0 Test  7>  The device is not online
MIC 0 Test  8 Check for normal mode                 : SKIPPED
MIC 0 Test  8> Prerequisite 'Ensure the device is online' failed:
MIC 0 Test  8>  The device is not online
MIC 0 Test  9 Check the POST code via SCIF          : SKIPPED
MIC 0 Test  9> Prerequisite 'Ensure the device is online' failed:
MIC 0 Test  9>  The device is not online
MIC 0 Test 10 Send data to the device               : SKIPPED
MIC 0 Test 10> Prerequisite 'Check for normal mode' failed:
MIC 0 Test 10>  The device is not in normal mode
MIC 0 Test 11 Compare the PCI configuration         : OK
MIC 0 Test 12 Ensure Flash version matches manifest : SKIPPED
MIC 0 Test 12> Prerequisite 'Check for normal mode' failed:
MIC 0 Test 12>  The device is not in normal mode
MIC 0 Test 13 Ping the host                         : SKIPPED
MIC 0 Test 13> Prerequisite 'Check for normal mode' failed:
MIC 0 Test 13>  The device is not in normal mode
Status: Error 0xc4: Service is not supported

The outpu of the /var/log/mpssd:

Tue Mar  5 23:12:20 2013: MPSS Daemon start
Tue Mar  5 23:12:20 2013: mic0: Hostname numtest3-mic0
Tue Mar  5 23:12:20 2013: mic0: MIC MAC address: 72:3f:d1:02:88:fd
Tue Mar  5 23:12:20 2013: mic0: Host MAC address: ee:83:17:f4:be:0f
Tue Mar  5 23:12:20 2013: mic0: Network static pair (Host 172.31.1.254 Card 172.31.1.1)
Tue Mar  5 23:12:20 2013: mic0: Configure host MAC address ee:83:17:f4:be:0f
Tue Mar  5 23:12:20 2013: mic0: Configure host IP address 172.31.1.254
Tue Mar  5 23:12:20 2013: mic0: Linux OS image /lib/firmware/mic/uos.img
Tue Mar  5 23:12:20 2013: mic0: Boot On Start Enabled
Tue Mar  5 23:12:20 2013: mic0: Shutdown Timeout 300
Tue Mar  5 23:12:20 2013: mic0: Crash Dump Limit in GB 16
Tue Mar  5 23:12:20 2013: mic0: MIC Crash Dump Directory at /var/crash/mic/
Tue Mar  5 23:12:20 2013: mic0: log_buf_addr: ffffffff83cb12d0
Tue Mar  5 23:12:20 2013: mic0: log_buf_len: ffffffff817249b0
Tue Mar  5 23:12:20 2013: mic0: Verbose mode Disabled
Tue Mar  5 23:12:20 2013: mic0: Using ramfs for root device
Tue Mar  5 23:12:20 2013: mic0: Base of file system at /opt/intel/mic/filesystem/base
Tue Mar  5 23:12:20 2013: mic0: Common files at /opt/intel/mic/filesystem/common
Tue Mar  5 23:12:20 2013: mic0: Unique files at /opt/intel/mic/filesystem/mic0
Tue Mar  5 23:12:20 2013: mic0: Overlay files at /opt/intel/mic/coi
Tue Mar  5 23:12:20 2013: mic0: Console hvc0
Tue Mar  5 23:12:20 2013: mic0: PowerManagement cpufreq_on;corec6_off;pc3_on;pc6_on
Tue Mar  5 23:12:20 2013: mic0: Command line: "quiet root=ramfs console=hvc0  micpm=cpufreq_on;corec6_off;pc3_on;pc6_on"
Tue Mar  5 23:12:20 2013: mic0: Booting /lib/firmware/mic/uos.img
Tue Mar  5 23:12:20 2013: mic0: State ready -> booting
Tue Mar  5 23:12:22 2013: Wait for download requests
Tue Mar  5 23:13:05 2013: Wait for card request with scif_accept() failed: Device or resource busy

[Frances_R_Intel]

The information that was passed to me on the initial problem was that the two most likely causes of this "SPI flash image ... is corrupted" error are: 1) the image really is corrupted or 2) the device driver was passing bad information to the micflash command. 

If I understand your reply, when you uninstalled and then reinstalled the MPSS, the flash update appeared to succeed. That would seem to imply that the image was not really corrupted. But it still looks suspiciously like the driver is not communicating as expected. 

Could you confirm for me that when you uninstalled the MPSS, you had first done a 

sudo service ofed-mic stop
sudo service mpss stop
sudo service mpss unload
sudo service opensmd stop
sudo service openibd stop

Also, could you confirm that you rebooted the host after reinstalling the MPSS and updating the flash?

[hpcdkhab]

I confirm that I had done:

sudo service mpss stop
sudo service mpss unload

I hadn't done

sudo service ofed-mic stop

sudo service opensmd stop

sudo service openibd stop

Unfortunately it was not in the readme file, which is in the KNC_gold_update_1-2.1.4982-15-rhel-6.3.

I have rebooted the Workstation in every step of installation.

But I could solve the problem.

The new installation and flashupdate brought nothing: KNC could not boot.
What has helped:
1) uninstall version of Gold
2) Installation of Alpha version and flashupdate with old Image
3) Installation of gold version

micinfo:

MicInfo Utility Log

Created Wed Mar  6 17:30:44 2013


    System Info
        Host OS                  : Linux
        OS Version               : 2.6.32-279.19.1.el6.x86_64
        Driver Version           : 4982-15
        MPSS Version             : 2.1.4982-15
        Host Physical Memory     : 24543 MB
        CPU Family               : GenuineIntel  Family 6  Model 44  Stepping 2
        CPU Speed                : 1596.000
        Threads per Core         : 2


Device No: 0,  Device Name: Intel(R) Xeon Phi(TM) Coprocessor

    Version
        Flash Version            : 2.1.05.0375
        UOS Version              : 2.6.38.8-g32944d0
        Device Serial Number     : ADKC22000467

    Board
        Vendor ID                : 8086
        Device ID                : 225c
        SubSystem ID             : 2500
        Coprocessor Stepping ID  : 1
        PCIe Width               : Insufficient Privileges
        PCIe Speed               : Insufficient Privileges
        PCIe Max payload size    : Insufficient Privileges
        PCIe Max read req size   : Insufficient Privileges
        Coprocessor Model        : 0x01
        Coprocessor Model Ext    : 0x00
        Coprocessor Type         : 0x00
        Coprocessor Family       : 0x0b
        Coprocessor Family Ext   : 0x00
        Coprocessor Stepping     : B0
        Board SKU                : ES2-P1750
        ECC Mode                 : Enabled
        SMC HW Revision          : Product 300W Active CS

    Core
        Total No of Active Cores : 61
        Voltage                  : 942000 uV
        Frequency                : 1090909 kHz

    Thermal
        Fan Speed Control        : On
        SMC Firmware Version     : 1.7.4172
        FSC Strap                : 14 MHz
        Fan RPM                  : 1200
        Fan PWM                  : 20
        Die Temp                 : 53 C

    GDDR
        GDDR Vendor              : Elpida
        GDDR Version             : 0x1
        GDDR Density             : 2048 Mb
        GDDR Size                : 7936 MB
        GDDR Technology          : GDDR5
        GDDR Speed               : 5.500000 GT/s
        GDDR Frequency           : 2750000 kHz
        GDDR Voltage             : 1000000 uV

Thank you for quick help

Dmitry