Conference rdgeng::cics_technical

189.0. "core generated by cicsterm" by NNTPD::"[email protected]" (Ricardo Lopez Cencerrado) Wed May 07 1997 23:04

Hello,

	I have obtained a core file generated by cicsterm on a system being
configured
for a customer. the core happens whenever calling cicsterm with no parameters
from inside a shell generated by the script command. In this situation, when
an available region is selected pressing return, cicsterm generates a core
file
and a cma_dump.log file is created in the current working directory.

	The contents of the cma_dump.log file are:

<<<< 
# cat cma_dump.log
%DECthreads bugcheck (version V3.12-312), terminating execution.
% Running on DEC OSF/1 AXP [OSF1 alpha V3.2(62); cpu type 39, configured for
14
%  cpus, 2 cpus in box, 1023Mb]
% Reason: test and set: high order bits corrupt at 0x63ce1a30
%     
%     The DECthreads library has detected an inconsistency in its internal
%   state and cannot continue execution. The inconsistency may be due to a bug
%   within the DECthreads library, the application program, or in any library
%   active in the address space. Common causes are unchecked stack overflows,
%   writes through uninitialized pointers, and synchronization races that
%   result in use of invalid data by some thread.
%     Application and library developers are requested to please check for
%   such problems before reporting a DECthreads library problem.
%     The information in this file may aid application program, library, or
%   DECthreads developers in determining the state of the process at the time
%   of this bugcheck. When the problem is reported, this information should be
%   included, along with a detailed procedure for reproducing the problem, if
%   that is possible. The 'detailed procedure' most likely to be of use to
%   developers is a complete program.
% 
% The bugcheck occurred at 
% ***CONCURRENT BUGCHECK***
% ***CANNOT CONTINUE: REPORTED STATE MAY BE INACCURATE AND INCOMPLETE***
>>>>

	The sequence of commands to produce the core is:

<<<<
capsu001 # script cicsterm_core.txt
Script started on Tue May  6 11:46:14 MET DST 1997
capsu001 # set -o emacs
capsu001 # env
MANPATH=/usr/i18n/share/%L/man:/usr/i18n/share/man:/usr/share/man:/usr/local/m
n
LANG=en_US.ISO8859-1
NLSPATH=:/usr/lib/nls/msg/%L/%N:/usr/lib/nls/msg/en_US.ISO8859-1/%N:/opt/encin
/msg/%L/%N
ENCINA_BINDING_FILE=/var/cics_servers/server_bindings
I18NPATH=/usr/i18n
PATH=/opt/cics/bin:/opt/encina/bin:/sbin:/usr/sbin:/usr/bin:/usr/ccs/bin:/usr/
in/X11:/usr/local
EDITOR=vi
LD_BIND_NOW=1
CICS_SFS_VG=ase_dev
LOGNAME=root
COBDIR=/usr/lib/cmplrs/cob
PS1=capsu001 # 
_RLD_ARGS=-ignore_unresolved
USER=root
SHELL=/bin/ksh
CICS_HOSTS=capsu001
HOME=/
LD_LIBRARY_PATH=/opt/cics/lib:/opt/encina/lib:/usr/lib/cmplrs/cob/coblib:/opt/
celocal/shlib
TERM=vt220
ENCINA_ROOT=/opt/encina
PWD=/scripts
DECW_BOOKSHELF=/opt/cics/dxbook/cics.decw_bookshelf
ENCINA_ROOT=/opt/encina
PWD=/scripts
DECW_BOOKSHELF=/opt/cics/dxbook/cics.decw_bookshelf
capsu001 # cicsterm 
vt220 keys
ERZ011060I/3710: Searching for available regions.  Please wait.
ERZ011102I/4321: Searching for regions on 'capsu001'
......
ERZ011062I/3714: Connecting to region 'CICS_TI'.
%DECthreads bugcheck (version V3.12-312), terminating execution.
% Running on DEC OSF/1 AXP [OSF1 alpha V3.2(62); cpu type 39, configured for
14
%  cpus, 2 cpus in box, 1023Mb]
% Reason: test and set: high order bits corrupt at 0x63ce1a30
% See 'cma_dump.log' for state information.
Resources lost(coredump)
>>>>

The environment file of the region is:

<<<<
ORACLE_HOME=/usr/oracle
ORACLE_SID=AIFOPEN_TU
>>>>

The stanza file of the region is:

<<<<
:ml5
ResourceDescription="Region Definition"
AmendCounter=1
Modifiable=no
StartType=cold
Groups=
StartupProgList=""
ShutdownProgList1=""
ShutdownProgList2=""
DefaultUserId="CICSUSER"
FileSystemType=SFS
RDBMSInstance=""
FileRSLCheck=external
TransientDataRSLCheck=external
TemporaryStorageRSLCheck=external
JournalRSLCheck=external
ProgramRSLCheck=external
TransactionRSLCheck=external
ESMLoad=no
ESMModule=""
RuntimeProtection=none
LogicalTDQProtection=none
PhysicalTDQProtection=none
NonRecTDQProtection=none
RecTSQProtection=none
NonRecTSQProtection=none
LocalQProtectProtection=none
LocalQProtection=none
ReleaseNum="0211"
LocalSysId="ISC0"
LocalNetworkName=""
CWASize=512
MinServer=1
MaxServer=5
ClassMaxTasks=1,1,1,1,1,1,1,1,1,1
ClassMaxTaskLim=0,0,0,0,0,0,0,0,0,0
ServerIdleLimit=300
SafetyLevel=normal
RPCListenerThreads=0
DateForm=ddmmyy
ClassTableSize=5,50,50,1,50,50,50,20,5,50,1,20,5,5
MaxRegionPool=2097152
MaxTaskPrivatePool=1048576
MaxTSHPool=1048576
RegionPoolThreshold=90
TaskShPoolThreshold=90
TaskSHNumBuckets=512
LoadDataNumBuckets=512
SysDump=no
PCDump=yes
ABDump=yes
DumpName="dumps"
CoreDumpName="dir1"
Trace=all
TraceModules=0
ExternalTrace=no
TraceFileA="trace.a"
TraceFileB="trace.b"
TraceFileSize=3276800
SysTraceBufferSize=163840
UserTraceDirectory="/tmp"
PublicUserTraceFile="cicspubl"
IntrospectInterval=10
IntrospectLevel=minimal
ISCDelayMinutes=10
CUBSDelayMinutes=5
CARPDelayHours=8
ProtectPurgeDelayPeriod=8
PurgeDelayPeriod=8
StatsRecord=yes
StatFile="statsfile"
SufficesSupported=yes
CheckpointInterval=1000
DefaultFileServer="/.:/cics/sfs/devel"
RecTSQFile="%Rcicsrectsqfile"
RecTSQIndex="cicsrectsqidx"
RecTSQVol="sfs_%S"
RecTSQPrePages=5
RecTSQMaxRecs=1000000
NonRecTSQFile="%Rcicsnrectsqfil"
NonRecTSQIndex="cicsnrectsqidx"
NonRecTSQVol="sfs_%S"
NonRecTSQPrePages=5
NonRecTSQMaxRecs=1000000
LogicalTDQFile="%Rcicstdqlgfile"
LogicalTDQIndex="cicstdqlgidx"
LogicalTDQVol="sfs_%S"
LogicalTDQPrePages=5
LogicalTDQMaxRecs=1000000
PhysicalTDQFile="%Rcicstdqphfile"
PhysicalTDQIndex="cicstdqphidx"
PhysicalTDQVol="sfs_%S"
PhysicalTDQPrePages=5
PhysicalTDQMaxRecs=1000000
NonRecTDQFile="%Rcicstdqnofile"
NonRecTDQIndex="cicstdqnoidx"
NonRecTDQVol="sfs_%S"
NonRecTDQPrePages=5
NonRecTDQMaxRecs=1000000
PhysicalTDQMaxRecs=1000000
NonRecTDQFile="%Rcicstdqnofile"
NonRecTDQIndex="cicstdqnoidx"
NonRecTDQVol="sfs_%S"
NonRecTDQPrePages=5
NonRecTDQMaxRecs=1000000
LocalQProtectFile="%Rcicsplqfile"
LocalQProtectIndex="cicsplqidx"
LocalQProtectVol="sfs_%S"
LocalQProtectPrePages=5
LocalQProtectMaxRecs=1000000
LocalQFile="%Rcicsnlqfile"
LocalQIndex="cicsnlqidx"
LocalQVol="sfs_%S"
LocalQPrePages=5
LocalQMaxRecs=1000000
TSQAgeLimit=20
ProgramCacheSize=0
LocalLUName=""
ServerSideTran=no
AuthenticationService=CICS
NameService=NONE
AllowDebugging=no
ECIPasswordCacheSeconds=28800
>>>>


This problem happens whenever using cicsterm with no parameters inside an
"script" shell. I could also post the corresponding cicsterm core file if
required. The region console.msg file shows no messages and no symrec appears.

The system is an 8400 5/440 with, 2 CPUs, 1 GB RAM, running OSF 3.2G, DCE 1.3B

runtime services and CICS 2.1A. the regions are empty, as when created and the

environment file for each region only contain ORACLE_HOME and ORACLE_SID
values.
Oracle version is 7.1.6.

Is there anything wrong with environment variables or the cics configuration ?


Any suggestion about solving this problem is wellcome.


Thanks in advance,


Ricardo Lopez Cencerrado.

[Posted by WWW Notes gateway]

Ricardo,

A DECthreads bugcheck with the reason you're seeing is normally due
to some code walking on the test and set bits belonging to DECthreads.
The code doing the walking may well not be the code that then sees the
bug check problem.

I've just tried reproducing your problem with the same CICS kit on
Digital UNIX V3.2D and I can't reproduce it.  I've tried a couple of
different scenarios but no luck.

The following is a list of things which are probably worth trying and
may help to narrow down the problem.

Is the region running on the same system as the cicsterm?  If it is, try
unsetting the CICS_HOST environment variable.

Is the system SMP?  If so have you tried the same thing on a single CPU
box.  Note that unfortunately I can't test this at the right levels on
an SMP box at the moment.

When you run cicsterm without any arguments outside of script, are you
using exactly the same environment?

Have you any of the Digital UNIX V3.2G patches installed - there's a new
patch kit which has gone up in the last couple of days.  I notice that
there is a patch for a memory corruption problem when running with Oracle
and Unicenter from Computer Associates (Patch OSF350-248).  This is a long
shot, but possible.

If you have a second region on the system, does the problem still occur?

Sorry that I can't give you a more precise answer on this one.  Please
post how you get on.

Regards,

Helen.

>> The region console.msg file shows no messages and no symrec appears.
Have you checked the file CSMT.out in the regions data directory?

>>Is there anything wrong with environment variables or the cics configuration ?
I can't see anything obvious.