IBM Support

IV69918: CRASH IN MLXENTDD DURING SEA FAILOVER APPLIES TO AIX 7100-03

A fix is available

Subscribe

You can track all active APARs for this component.

APAR status

  • Closed as program error.

Error description

  • A DSI will occur with following stack whenever the BACKUP
    SEA configured over ROCE port is rebooted.
    
    [04731644]mlxentdd:tmetrics_delta+000024
    (00000000000F667E, 00000000000001B0 [??])
    [04753FBC]mlxentdd:rdmacore_query_port+00075C (??, ??,
    ??)
    [047742F0]mlxentdd:mlxent_hw_rdma_query_port+000010 (??,
    ??, ??)
    [04F91800]ofed_core:ib_query_port+0000C0 (??, ??, ??)
    [04F8FD54]ofed_core:read_port_table_lengths+000114 (??)
    [04F930EC]ofed_core:ib_register_device+0002EC (??)
    [004CE3D4]ofed_register_device+000054 (??)
    [00014F50].kernel_add_gate_cstack+000030 ()
    [046F3FD4]mlxentdd:rdmacore_callback_ofed+0000D4 (??)
    [004CDFB0]ofed_call_notifiers+000030 ()
    [004CDF24]ofed_register_verbs+000024 (??)
    [00014F50].kernel_add_gate_cstack+000030 ()
    [04F82E24]ofed_core:ofed_init+0001E4 ()
    [04F80778]ofed_core:ofed_dd_config+000338 (??, ??, ??)
    [005D77DC]config_dd+00029C (??, ??, ??)
    [005D8600]sysconfig+000240 (??, ??, ??)
    
    This typically will happen whenever SEA on VIOS is
    transitioning to BACKUP mode.
    

Local fix

  • Customer reports that doing:
    
    chdev -l entN -a adapter_reset=no
    
    where entN is the SEA tends to prevent the DSI.
    

Problem summary

  • System crash when VIOS with BACKUP SEA configured over ROCE
    port is rebooted AND rdma attribute set to "desired" or
    "required".
    

Problem conclusion

  • Code changed to yield RDMA whenever SEA is configured over ROCE
    port with rdma attribute set to "desired" or "required".
    

Temporary fix

Comments

  • 6100-08 - use AIX APAR IV69976
    6100-09 - use AIX APAR IV56582
    6100-09 - use AIX APAR IV56582
    7100-02 - use AIX APAR IV70232
    7100-03 - use AIX APAR IV69918
    

APAR Information

  • APAR number

    IV69918

  • Reported component name

    AIX V7.1

  • Reported component ID

    5765H4000

  • Reported release

    710

  • Status

    CLOSED PER

  • PE

    NoPE

  • HIPER

    NoHIPER

  • Submitted date

    2015-02-25

  • Closed date

    2015-02-25

  • Last modified date

    2015-09-30

  • APAR is sysrouted FROM one or more of the following:

    IV56582

  • APAR is sysrouted TO one or more of the following:

Fix information

  • Fixed component name

    AIX V7.1

  • Fixed component ID

    5765H4000

Applicable component levels

  • R710 PSY U865812

       UP15/05/19 I 1000

PTF to Fileset Mapping



Document information

More support for: AIX Enterprise Edition

Software version: 710

Operating system(s): AIX

Reference #: IV69918

Modified date: 30 September 2015