hello managers,
this was a memory error in some cache, which was corrected by the
ECC hardware on the board. there seems to be no problem, as long
such messages do not occure often, which they currently don't.
in case i see this more often, i could try to:
* reboot. maybe the problem simply goes away.
* reseat the simms.
* call deq for service. they will probably replace simms.
many thanks to:
   Joe Fletcher <joe_at_meng.ucl.ac.uk>
   Simon Millard <Simon.Millard_at_barclays.co.uk>
   Chris Wilson <Chris.Wilson_at_acco-uk.co.uk>
for their really quick replies and
   alan_at_nabeth.cxo.dec.com
for a more detailed description.
the original message:
< hallo managers,
< 
< on a DEC 3000 running Tru64 4.0D (without patches) with 64 MB ram
< I found this in /var/adm/messages:
< 
< Jun 16 18:35:26 oek vmunix: Memory error corrected by processor
< Jun 16 18:35:27 oek vmunix:       biu_stat        = 0000000000003340
< Jun 16 18:35:27 oek vmunix:       biu_addr        = 0000000000108018
< Jun 16 18:35:27 oek vmunix:       dc_stat         = 0000000000000007
< Jun 16 18:35:27 oek vmunix:       fill_syndrome   = 0000000000003680
< Jun 16 18:35:27 oek vmunix:       fill_addr       = 0000000000fe3a40
< Jun 16 18:35:27 oek vmunix:       bc_tag          = 0000000000400f95
< Jun 16 18:35:27 oek vmunix:       ident           = 0
< 
< uerf says:
< 
< <----- EVENT INFORMATION -----
< <
< <EVENT CLASS                             ERROR EVENT 
< <OS EVENT TYPE                  100.     CPU EXCEPTION 
< <SEQUENCE NUMBER                  1.
< <OPERATING SYSTEM                        DEC OSF/1 
< <OCCURRED/LOGGED ON                      Wed Jun 16 18:35:26 1999
< <OCCURRED ON SYSTEM                      oek 
< <SYSTEM ID                 x00020004     CPU TYPE:  DEC 3000 
< <SYSTYPE                   x00000000
< <
< <----- UNIT INFORMATION -----
< <
< <UNIT CLASS                              CPU 
< <
< <----- KN15AA CPU 630/620 STACK FRAME -----
< <
< <PROCESSOR OFFSET          x00000018
< <SYSTEM OFFSET             x00000048
< <BIU_STAT              x0000000000003340
< <                                    BIU_CMD CYCLE CLASS IS READ_BLOCK 
< <                                     FILL_ECC PRI. CACHE FILL FROM EXT. 
< <                                      _CACHE HAD ECC ERROR 
< <BIU_ADDR              x0000000000108018
< <                                     PHYSICAL ADDRESS OF CACHE BLOCK
< <WITH ERROR IS x8400
< <DC_STAT               x0000000000000007
< <                                     DC_HIT LAST LOAD OR STORE MISSED 
< <                                      _DCACHE 
< <                                     OPCODE RA FIELD - INTEGER REGISTER
< <IS R 0.
< <FILL_SYNDROME         x0000000000003680  SINGLE BIT ERROR IS NO ERRORS 
< <                                     SINGLE BIT ERROR IS DATA BIT 29 
< <FILL_ADDR             x0000000000FE3A40
< <                                     PHYSICAL ADDRESS OF QUADWORD WITH
< <ERROR x7F1D2
< <BC_TAG                x0000000000400F95  EXTERNAL CACHE TAG CONTROL
< <BITS 
< <                                      _EXTERNAL CACHE HIT 
< <                                     D BIT - CACHE BLOCK DIRTY 
< <                                     V BIT - CACHE BLOCK VALID 
< <                                     TAG ADDRESS IS x7C
< <                                     EXTERNAL CACHE TAG CONTROL BITS
< <TAG 
< <                                      _ADDRESS PARITY BIT 
< <INT_EXC_IDENT         x0000000000000000
< <                                     INTERRUPT OR EXCEPTION IS NONE 
< 
< Any idea what this means and what problems might folow?
< the machine is running fine otherwise.
< 
< thanx in advance,
< andreas
yours truly,
andreas
--
 __  ___ __   Andreas Bunten -=- BOFH_at_unixmail.pz-oekosys.uni-kiel.de _
/__`|__ / _`  If you think nobody cares if you're alive, try       ,_(')<
.__/|___\__>  missing a couple of car payments. -- Earl Wilson     \___)
Received on Mon Jun 28 1999 - 18:28:24 NZST