Where have the control blocks gone?
Steve Comstock
predecessors since 1975, and feel I have a pretty
good idea how to use a SYSUDUMP. But things are
getting a little mysterious right now.
I'm getting lots of dumps as I create and debug
a new application, and suddenly I've noticed:
* Instead of three (or more) SVRBs, there are only two
* When I force an abend right after OPEN the DCB does
not show up as formatted in the dump (I can find the
storage location and figure it out, but I'm surprised
at this development).
Have there been some changes in SYSUDUMP formatting in
recent years? (I'm running z/OS 1.10 currently).
--
Kind regards,
-Steve Comstock
The Trainer's Friend, Inc.
303-393-8716
http://www.trainersfriend.com
z/OS Application development made easier
* Our classes include
+ How things work
+ Programming examples with realistic applications
+ Starter / skeleton code
+ Complete working programs
+ Useful utilities and subroutines
+ Tips and techniques
==> Ask about being added to our opt-in list: <==
==> * Early announcement of new courses <==
==> * Early announcement of new techincal papers <==
==> * Early announcement of new promotions <==
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to list...@bama.ua.edu with the message: GET IBM-MAIN INFO
Search the archives at http://bama.ua.edu/archives/ibm-main.html
Thompson, Steve
From: IBM Mainframe Discussion List [mailto:IBM-...@bama.ua.edu] On
Behalf Of Steve Comstock
Sent: Monday, January 04, 2010 11:38 AM
To: IBM-...@bama.ua.edu
Subject: Where have the control blocks gone?
I've been debugging applications in z/OS and its
predecessors since 1975, and feel I have a pretty
good idea how to use a SYSUDUMP. But things are
getting a little mysterious right now.
I'm getting lots of dumps as I create and debug
a new application, and suddenly I've noticed:
* Instead of three (or more) SVRBs, there are only two
* When I force an abend right after OPEN the DCB does
not show up as formatted in the dump (I can find the
storage location and figure it out, but I'm surprised
at this development).
Have there been some changes in SYSUDUMP formatting in
recent years? (I'm running z/OS 1.10 currently).
<SNIP>
Check PARMLIB settings for dumps. Someone may have changed them on you.
Regards,
Steve Thompson
Mike Myers
I have been debugging dumps since OS/360 days (around 1967). While it is
possible to have three SVRBs as the last active elements in the RB
queue, it is far more typical in the case of an ABEND in an application
program to have only two SVRBs. This assumes that the application
program is running under a PRB (as is typical). When it ABENDs, an SVRB
is created under which the ABEND process (SVC D) will run, followed by a
second SVRB under which SNAP (SVC 33x - 51 decimal) runs to produce the
dump. In the case you cite, where there are three SVRBs, that would
occur if the application (running under a PRB) were to call a system
service (such as OPEN) and caused it to ABEND by passing bad parameters.
While this latter case is certainly possible, it is usually the former
case which is most likely to occur for an application ABEND.
As for the DCB, I don't ever recall it being formatted in a dump, except
possibly one produced by ABENDAID. It may not have even been formatted
there either, as I personally had little contact with ABENDAID. I can't
say I recall one being formatted in a SYSUDUMP. While it may have been
present in the dump, I can only recall ever seeing it laid out as a
block of data, but not formatted into firlds.
I do know that the formatter in IPCS (TCBEXIT IECDAFMT - at least I
think that's the right exit - if not, then it's probably IECIOFMT) will
display the DCB as a block of data, but does not format its fields either.
Mike Myers
Mentor Services Corporation
Steve Comstock wrote:
> I've been debugging applications in z/OS and its
> predecessors since 1975, and feel I have a pretty
> good idea how to use a SYSUDUMP. But things are
> getting a little mysterious right now.
>
> I'm getting lots of dumps as I create and debug
> a new application, and suddenly I've noticed:
>
> * Instead of three (or more) SVRBs, there are only two
>
> * When I force an abend right after OPEN the DCB does
> not show up as formatted in the dump (I can find the
> storage location and figure it out, but I'm surprised
> at this development).
>
>
> Have there been some changes in SYSUDUMP formatting in
> recent years? (I'm running z/OS 1.10 currently).
>
>
----------------------------------------------------------------------
Steve Comstock
> -----Original Message-----
> From: IBM Mainframe Discussion List [mailto:IBM-...@bama.ua.edu] On
> Behalf Of Steve Comstock
> Sent: Monday, January 04, 2010 11:38 AM
> To: IBM-...@bama.ua.edu
> Subject: Where have the control blocks gone?
>
> I've been debugging applications in z/OS and its
> predecessors since 1975, and feel I have a pretty
> good idea how to use a SYSUDUMP. But things are
> getting a little mysterious right now.
>
> I'm getting lots of dumps as I create and debug
> a new application, and suddenly I've noticed:
>
> * Instead of three (or more) SVRBs, there are only two
>
> * When I force an abend right after OPEN the DCB does
> not show up as formatted in the dump (I can find the
> storage location and figure it out, but I'm surprised
> at this development).
>
>
> Have there been some changes in SYSUDUMP formatting in
> recent years? (I'm running z/OS 1.10 currently).
>
>
> <SNIP>
>
> Check PARMLIB settings for dumps. Someone may have changed them on you.
>
> Regards,
> Steve Thompson
Nope.
--
Kind regards,
-Steve Comstock
The Trainer's Friend, Inc.
303-393-8716
http://www.trainersfriend.com
z/OS Application development made easier
* Our classes include
+ How things work
+ Programming examples with realistic applications
+ Starter / skeleton code
+ Complete working programs
+ Useful utilities and subroutines
+ Tips and techniques
==> Ask about being added to our opt-in list: <==
==> * Early announcement of new courses <==
==> * Early announcement of new techincal papers <==
==> * Early announcement of new promotions <==
----------------------------------------------------------------------
Steve Comstock
> Steve:
>
> I have been debugging dumps since OS/360 days (around 1967). While it is
> possible to have three SVRBs as the last active elements in the RB
> queue, it is far more typical in the case of an ABEND in an application
> program to have only two SVRBs. This assumes that the application
> program is running under a PRB (as is typical). When it ABENDs, an SVRB
> is created under which the ABEND process (SVC D) will run, followed by a
> second SVRB under which SNAP (SVC 33x - 51 decimal) runs to produce the
> dump. In the case you cite, where there are three SVRBs, that would
> occur if the application (running under a PRB) were to call a system
> service (such as OPEN) and caused it to ABEND by passing bad parameters.
> While this latter case is certainly possible, it is usually the former
> case which is most likely to occur for an application ABEND.
Right. Went back to my notes and that's what I have; should
have checked first. My mantra has been "third RB from the
bottom is the villain" and I knew it could be (and usually is)
a PRB, but got tangled up in recall as "third SVRB".
>
> As for the DCB, I don't ever recall it being formatted in a dump, except
> possibly one produced by ABENDAID. It may not have even been formatted
> there either, as I personally had little contact with ABENDAID. I can't
> say I recall one being formatted in a SYSUDUMP. While it may have been
> present in the dump, I can only recall ever seeing it laid out as a
> block of data, but not formatted into firlds.
Ah, the ever-wonderful firlds. :-)
Anyway, you're right on this count, too. Guess I've been
concentrating too deeply on one area and not paying
attention to what I already know but mis-remembered.
Thanks for your note. I'll go back to my code now.
>
> I do know that the formatter in IPCS (TCBEXIT IECDAFMT - at least I
> think that's the right exit - if not, then it's probably IECIOFMT) will
> display the DCB as a block of data, but does not format its fields either.
>
> Mike Myers
> Mentor Services Corporation
>
> Steve Comstock wrote:
>> I've been debugging applications in z/OS and its
>> predecessors since 1975, and feel I have a pretty
>> good idea how to use a SYSUDUMP. But things are
>> getting a little mysterious right now.
>>
>> I'm getting lots of dumps as I create and debug
>> a new application, and suddenly I've noticed:
>>
>> * Instead of three (or more) SVRBs, there are only two
>>
>> * When I force an abend right after OPEN the DCB does
>> not show up as formatted in the dump (I can find the
>> storage location and figure it out, but I'm surprised
>> at this development).
>>
>>
>> Have there been some changes in SYSUDUMP formatting in
>> recent years? (I'm running z/OS 1.10 currently).
>>
>>
>
> ----------------------------------------------------------------------
> For IBM-MAIN subscribe / signoff / archive access instructions,
> send email to list...@bama.ua.edu with the message: GET IBM-MAIN INFO
> Search the archives at http://bama.ua.edu/archives/ibm-main.html
>
--
Kind regards,
-Steve Comstock
The Trainer's Friend, Inc.
303-393-8716
http://www.trainersfriend.com
z/OS Application development made easier
* Our classes include
+ How things work
+ Programming examples with realistic applications
+ Starter / skeleton code
+ Complete working programs
+ Useful utilities and subroutines
+ Tips and techniques
==> Ask about being added to our opt-in list: <==
==> * Early announcement of new courses <==
==> * Early announcement of new techincal papers <==
==> * Early announcement of new promotions <==
----------------------------------------------------------------------
Thompson, Steve
From: IBM Mainframe Discussion List [mailto:IBM-...@bama.ua.edu] On
Behalf Of Steve Comstock
Sent: Monday, January 04, 2010 4:27 PM
To: IBM-...@bama.ua.edu
Subject: Re: Where have the control blocks gone?
Thompson, Steve wrote:
> -----Original Message-----
> From: IBM Mainframe Discussion List [mailto:IBM-...@bama.ua.edu] On
> Behalf Of Steve Comstock
> Sent: Monday, January 04, 2010 11:38 AM
> To: IBM-...@bama.ua.edu
> Subject: Where have the control blocks gone?
>
> I've been debugging applications in z/OS and its
> predecessors since 1975, and feel I have a pretty
> good idea how to use a SYSUDUMP. But things are
> getting a little mysterious right now.
>
> I'm getting lots of dumps as I create and debug
> a new application, and suddenly I've noticed:
>
> * Instead of three (or more) SVRBs, there are only two
>
> * When I force an abend right after OPEN the DCB does
> not show up as formatted in the dump (I can find the
> storage location and figure it out, but I'm surprised
> at this development).
>
>
> Have there been some changes in SYSUDUMP formatting in
> recent years? (I'm running z/OS 1.10 currently).
>
>
> <SNIP>
>
> Check PARMLIB settings for dumps. Someone may have changed them on
you.
>
> Regards,
> Steve Thompson
Nope.
<SNIP>
Yeah, I was deep in thought on IPCS -- somehow the SYSUDUMP didn't
register right. And then I have to remember, we have written our own
formatting tools for various CBs...
Regards,
Steve Thompson
Mike Myers
Good mantra "third RB ...". I've used the same one many times when
teaching dump reading.
Mike
Don Poitras
>
> Steve:
>
>
> I do know that the formatter in IPCS (TCBEXIT IECDAFMT - at least I
> think that's the right exit - if not, then it's probably IECIOFMT) will
> display the DCB as a block of data, but does not format its fields either.
>
> Mike Myers
IP SUMM FORMAT will display DCB info, here's the one for SYSMDUMP:
DATA MANAGEMENT,CONTROL BLOCKS FOR OPEN DATA SETS
***DCB,AT LOCATION 7ADF68 (BSAM) DASD
+0 , DCBE 7F424B7C KEYCN 00 FDAD 000000 076F0005 02 DVTBA EF6026
+10 , KEYLE 00 DEVT 2F TRBAL 184E BUFCB 00000001 BUFL 0000 DSORG 4000 IOBAD 0050C000
+20 , EODAD 86000001 RECFM 98 EXLST 7ADFC4 TIOT 048C MACRF 0020 IFLGS 00 DEBA 7B516C
+30 , OFLGS 92 BEEA68 OPTCD 00 E00908 IOBL 0A SYNA 000001 CIND 0000 BLKSI 6180
+40 , WCPO 30013030 IOBA 0050C070 NCP 04 EOBRA D79658 EOBW 00D79658 DIRCT 0000 LRECL 1040
+54 , POINT 00000001
--
Don Poitras - zSeries R & D - SAS Institute Inc. - SAS Campus Drive
mailto:sas...@sas.com (919)531-5637 Fax:677-4444 Cary, NC 27513
Clark Morris
>Mike Myers wrote:
>>
>> Steve:
>>
>>
>> I do know that the formatter in IPCS (TCBEXIT IECDAFMT - at least I
>> think that's the right exit - if not, then it's probably IECIOFMT) will
>> display the DCB as a block of data, but does not format its fields either.
>>
>> Mike Myers
>
>IP SUMM FORMAT will display DCB info, here's the one for SYSMDUMP:
>
>DATA MANAGEMENT,CONTROL BLOCKS FOR OPEN DATA SETS
>***DCB,AT LOCATION 7ADF68 (BSAM) DASD
>+0 , DCBE 7F424B7C KEYCN 00 FDAD 000000 076F0005 02 DVTBA EF6026
>+10 , KEYLE 00 DEVT 2F TRBAL 184E BUFCB 00000001 BUFL 0000 DSORG 4000 IOBAD 0050C000
>+20 , EODAD 86000001 RECFM 98 EXLST 7ADFC4 TIOT 048C MACRF 0020 IFLGS 00 DEBA 7B516C
>+30 , OFLGS 92 BEEA68 OPTCD 00 E00908 IOBL 0A SYNA 000001 CIND 0000 BLKSI 6180
>+40 , WCPO 30013030 IOBA 0050C070 NCP 04 EOBRA D79658 EOBW 00D79658 DIRCT 0000 LRECL 1040
>+54 , POINT 00000001
The LE forms of SYSUDUMP (may be CEEDUMP, my memory is bad), have some
very useful formatting and make at least COBOL debugging much easier.
Don Poitras
>
> On 5 Jan 2010 07:17:48 -0800, in bit.listserv.ibm-main you wrote:
>
> >Mike Myers wrote:
> >>
> >> Steve:
> >>
> >>
> >> I do know that the formatter in IPCS (TCBEXIT IECDAFMT - at least I
> >> think that's the right exit - if not, then it's probably IECIOFMT) will
> >> display the DCB as a block of data, but does not format its fields either.
> >>
> >> Mike Myers
> >
> >IP SUMM FORMAT will display DCB info, here's the one for SYSMDUMP:
> >
> >DATA MANAGEMENT,CONTROL BLOCKS FOR OPEN DATA SETS
> >***DCB,AT LOCATION 7ADF68 (BSAM) DASD
> >+0 , DCBE 7F424B7C KEYCN 00 FDAD 000000 076F0005 02 DVTBA EF6026
> >+10 , KEYLE 00 DEVT 2F TRBAL 184E BUFCB 00000001 BUFL 0000 DSORG 4000 IOBAD 0050C000
> >+20 , EODAD 86000001 RECFM 98 EXLST 7ADFC4 TIOT 048C MACRF 0020 IFLGS 00 DEBA 7B516C
> >+30 , OFLGS 92 BEEA68 OPTCD 00 E00908 IOBL 0A SYNA 000001 CIND 0000 BLKSI 6180
> >+40 , WCPO 30013030 IOBA 0050C070 NCP 04 EOBRA D79658 EOBW 00D79658 DIRCT 0000 LRECL 1040
> >+54 , POINT 00000001
>
> The LE forms of SYSUDUMP (may be CEEDUMP, my memory is bad), have some
> very useful formatting and make at least COBOL debugging much easier.
I can't think of anything offhand that you can't get from:
IP VERBX LEDATA
--
Don Poitras - zSeries R & D - SAS Institute Inc. - SAS Campus Drive
mailto:sas...@sas.com (919)531-5637 Fax:677-4444 Cary, NC 27513
----------------------------------------------------------------------
Clark Morris
>Clark Morris wrote:
>>
>> On 5 Jan 2010 07:17:48 -0800, in bit.listserv.ibm-main you wrote:
>>
>> >Mike Myers wrote:
>> >>
>> >> Steve:
>> >>
>> >>
>> >> I do know that the formatter in IPCS (TCBEXIT IECDAFMT - at least I
>> >> think that's the right exit - if not, then it's probably IECIOFMT) will
>> >> display the DCB as a block of data, but does not format its fields either.
>> >>
>> >> Mike Myers
>> >
>> >IP SUMM FORMAT will display DCB info, here's the one for SYSMDUMP:
>> >
>> >DATA MANAGEMENT,CONTROL BLOCKS FOR OPEN DATA SETS
>> >***DCB,AT LOCATION 7ADF68 (BSAM) DASD
>> >+0 , DCBE 7F424B7C KEYCN 00 FDAD 000000 076F0005 02 DVTBA EF6026
>> >+10 , KEYLE 00 DEVT 2F TRBAL 184E BUFCB 00000001 BUFL 0000 DSORG 4000 IOBAD 0050C000
>> >+20 , EODAD 86000001 RECFM 98 EXLST 7ADFC4 TIOT 048C MACRF 0020 IFLGS 00 DEBA 7B516C
>> >+30 , OFLGS 92 BEEA68 OPTCD 00 E00908 IOBL 0A SYNA 000001 CIND 0000 BLKSI 6180
>> >+40 , WCPO 30013030 IOBA 0050C070 NCP 04 EOBRA D79658 EOBW 00D79658 DIRCT 0000 LRECL 1040
>> >+54 , POINT 00000001
>>
>> The LE forms of SYSUDUMP (may be CEEDUMP, my memory is bad), have some
>> very useful formatting and make at least COBOL debugging much easier.
>
>I can't think of anything offhand that you can't get from:
>
> IP VERBX LEDATA
Isn't that using SYSMDUMP and IPCS? I'm talking about the dumps
available to most applications programmers.
Don Poitras
There's nothing stopping any programmer changing his JCL from:
//SYSUDUMP to
//SYSMDUMP
There is a learning curve, but if looking at dumps is something you do
for a living, it's worth the effort. I avoided it for years because I
couldn't see the benefit, but now I just groan when somebody hands me a
huge SYSUDUMP.
--
Don Poitras - zSeries R & D - SAS Institute Inc. - SAS Campus Drive
mailto:sas...@sas.com (919)531-5637 Fax:677-4444 Cary, NC 27513
----------------------------------------------------------------------
Thompson, Steve
From: IBM Mainframe Discussion List [mailto:IBM-...@bama.ua.edu] On
Behalf Of Don Poitras
Sent: Wednesday, January 06, 2010 8:44 AM
To: IBM-...@bama.ua.edu
Subject: Re: Where have the control blocks gone?
<SNIPPAGE>
There's nothing stopping any programmer changing his JCL from:
//SYSUDUMP to
//SYSMDUMP
There is a learning curve, but if looking at dumps is something you do
for a living, it's worth the effort. I avoided it for years because I
couldn't see the benefit, but now I just groan when somebody hands me a
huge SYSUDUMP.
<SNIP>
I've been following this thread off and on. There is a caveat to the
SYSMDUMP that I think needs to be recognized (and if someone has a way
around it, I'm all eyes, ears or whatever).
SDUMP[X] knowledgeable programs, or programs with error recovery
routines: When you use SYSMDUMP, the last dump written clobbers what
went before.
That is the gotcha you have to be willing to accept. If the last dump
allows you to determine what the root cause is... <VBG>
Regards,
Steve Thompson
Farley, Peter x23353
Which may not be the case with COBOL internal sorts, where the SORT also
abends after the original abend occurs in an input or output procedure.
And not every site makes IPCS readily available to ordinary programmers,
even if they want to learn.
Peter
This message and any attachments are intended only for the use of the addressee and
may contain information that is privileged and confidential. If the reader of the
message is not the intended recipient or an authorized representative of the
intended recipient, you are hereby notified that any dissemination of this
communication is strictly prohibited. If you have received this communication in
error, please notify us immediately by e-mail and delete the message and any
attachments from your system.
Shmuel Metz , Seymour J.
at 10:37 AM, Steve Comstock <st...@TRAINERSFRIEND.COM> said:
>* Instead of three (or more) SVRBs, there are only two
Under what circumstances did you see three SVRB's in the past? I'd expect
that for an ABEND in O/C/EOV, but not for an ABEND in application code.
>* When I force an abend right after OPEN the DCB does
> not show up as formatted in the dump
The DEB should show up. AFAIK IBM never formatted the DCB in SYSUDUMP.
Were you thinking of SYSMDUMP viewed from IPCS?
--
Shmuel (Seymour J.) Metz, SysProg and JOAT
ISO position; see <http://patriot.net/~shmuel/resume/brief.html>
We don't care. We don't have to care, we're Congress.
(S877: The Shut up and Eat Your spam act of 2003)
Don Poitras
>
> -----Original Message-----
> From: IBM Mainframe Discussion List [mailto:IBM-...@bama.ua.edu] On
> Behalf Of Don Poitras
> Sent: Wednesday, January 06, 2010 8:44 AM
> To: IBM-...@bama.ua.edu
> Subject: Re: Where have the control blocks gone?
>
> <SNIPPAGE>
> There's nothing stopping any programmer changing his JCL from:
> //SYSUDUMP to
> //SYSMDUMP
>
> There is a learning curve, but if looking at dumps is something you do
> for a living, it's worth the effort. I avoided it for years because I
> couldn't see the benefit, but now I just groan when somebody hands me a
> huge SYSUDUMP.
> <SNIP>
>
> I've been following this thread off and on. There is a caveat to the
> SYSMDUMP that I think needs to be recognized (and if someone has a way
> around it, I'm all eyes, ears or whatever).
>
> SDUMP[X] knowledgeable programs, or programs with error recovery
> routines: When you use SYSMDUMP, the last dump written clobbers what
> went before.
>
> That is the gotcha you have to be willing to accept. If the last dump
> allows you to determine what the root cause is... <VBG>
Use FREE=CLOSE on the SYSMDUMP DD.
--
Don Poitras - zSeries R & D - SAS Institute Inc. - SAS Campus Drive
mailto:sas...@sas.com (919)531-5637 Fax:677-4444 Cary, NC 27513
----------------------------------------------------------------------
Jim Mulder
10:07:11 AM:
> > I do know that the formatter in IPCS (TCBEXIT IECDAFMT - at least I
> > think that's the right exit - if not, then it's probably IECIOFMT)
will
> > display the DCB as a block of data, but does not format its fields
either.
> >
> > Mike Myers
>
> IP SUMM FORMAT will display DCB info, here's the one for SYSMDUMP:
>
> DATA MANAGEMENT,CONTROL BLOCKS FOR OPEN DATA SETS
> ***DCB,AT LOCATION 7ADF68 (BSAM) DASD
> +0 , DCBE 7F424B7C KEYCN 00 FDAD 000000 076F0005 02
> DVTBA EF6026
> +10 , KEYLE 00 DEVT 2F TRBAL 184E BUFCB 00000001 BUFL 0000
> DSORG 4000 IOBAD 0050C000
> +20 , EODAD 86000001 RECFM 98 EXLST 7ADFC4 TIOT 048C
> MACRF 0020 IFLGS 00 DEBA 7B516C
> +30 , OFLGS 92 BEEA68 OPTCD 00 E00908 IOBL 0A
> SYNA 000001 CIND 0000 BLKSI 6180
> +40 , WCPO 30013030 IOBA 0050C070 NCP 04 EOBRA D79658
> EOBW 00D79658 DIRCT 0000 LRECL 1040
> +54 , POINT 00000001
Ownership of IECDAFMT was transferred from the BCP to DFSMS in
z/OS 1.11, and Wayne Rhoten enhanced its function considerably
in that release.
IECDAFMT is invoked under IPCS and also under SNAP (thus also
SYSABEND and SYSUDUMP), so the formatting enhancements should be visible
in both environments.
Jim Mulder z/OS System Test IBM Corp. Poughkeepsie, NY