John, you might be interested in this thread:
http://www.mail-archive.com/[email protected]/msg01761.html
Specifically, the binder option COMPAT=LKED.
LKED
Specifies that certain binder processing options are to work in a manner
compatible with the linkage editor. Specific processing affected by this
specification includes:
* REUS—If a section is encountered in a module with a lower reusability
than that specified on the REUS option, the reusability of the module is
automatically downgraded. An information message is issued and the return code
remains unchanged.
This may or not be useful, because in the end it seems to me that you DO want
to make the assembler module re-entrant. It wasn't clear to me if you had made
progress in doing so.
I will say that, though I put the project on hold after initial success with
this, I (a mere application developer) was able to relatively easily convert
some batch/CICS shared assembler routines to be re-entrant by LE enabling them.
I even wrote a macro (AUTOSTG) to help me with this.
CVASCBLK CSECT
CEEDSA ,
CEECAA ,
*
CVASCBLK CEEENTRY MAIN=NO,AUTO=L'CBLKAUTO
CBLKAUTO AUTOSTG USING=YES
MVC PARMLIST(PARMLGTH),0(R1) SAVE CALLING LIST
LM R4,R9,0(R1) LOAD PARAMETERS FROM PROGRAM
[...]
EXIT DS 0H
CEETERM RC=0
SPACE 3
LTORG ,
CBLKAUTO AUTOSTG DSECT=START
PARMLIST DS 0F
PARMINPM DS F
PARMOUTP DS F
PARMMSGL DS F
PARMBLKC DS F
PARMBLKS DS F
PARMOFFS DS F
PARMDELM DS F
PARMLGTH EQU *-PARMLIST
TRT DC XL256'00'
CBLKAUTO AUTOSTG DSECT=END
Original code
CVASCBLK CSECT
CVASCBLK AMODE ANY
CVASCBLK RMODE ANY
SAVE (14,12)
BALR R3,0
USING *,R3
B START
DC CL8'CVASCBLK'
START DS 0H
MVC PARMLIST(PARMLGTH),0(R1) SAVE CALLING LIST
ST R13,SAVE13
LA R13,CBLKSAVE
LM R4,R9,0(R1) LOAD PARAMETERS FROM PROGRAM
[...]
EXIT DS 0H
L R13,SAVE13
RETURN (14,12),RC=0
SPACE 3
LTORG ,
PARMLIST DS 0F
PARMINPM DS F
PARMOUTP DS F
PARMMSGL DS F
PARMBLKC DS F
PARMBLKS DS F
PARMOFFS DS F
PARMDELM DS F
PARMLGTH EQU *-PARMLIST
DS 0D
CBLKSAVE DS 9D
SAVE13 DS F
TRT DC XL256'00'
Obviously any non-reentrant code needs to be made reentrant.
And here is the AUTOSTG macro. (Wow, its a lot smaller than I'd remembered.)
.**********************************************************************
.* NOTE THAT CEEDSA MACRO REQUIRED PRIOR TO FIRST AUTOSTG MACRO *
.**********************************************************************
MACRO
&LABEL AUTOSTG &DSECT=,&USING=NO
GBLC &CEEOEPVNX Label of Main Entry; from CEEENTRY
LCLC &AUTOSECT
&AUTOSECT SETC '@&LABEL'
.DSECT ANOP ,
AIF ('&DSECT' EQ 'START').START
AIF ('&DSECT' EQ 'END').END
AGO .CHKUSING
MEXIT ,
.**********************************************************************
.* DSECT=START *
.**********************************************************************
.START ANOP ,
&AUTOSECT DSECT ,
ORG *+CEEDSASZ Leave space for the DSA fixed part
AGO .CHKUSING
.**********************************************************************
.* DSECT=END *
.**********************************************************************
.END ANOP ,
LCLC &OEPVNX
&LABEL EQU &AUTOSECT,*-&AUTOSECT,C'C'
&OEPVNX SETC '&CEEOEPVNX'
AIF ('&OEPVNX' EQ '').CHKUSING
&OEPVNX CSECT , Continue primary CSECT
.**********************************************************************
.CHKUSING AIF ('&USING' NE 'YES').EXIT
.**********************************************************************
.* USING=YES *
.**********************************************************************
USING &LABEL,CEEDSA Auto stg and DSA use the same base
.EXIT MEXIT
.**********************************************************************
MEND
EJECT
>________________________________
> From: John McKown <[email protected]>
>To: [email protected]
>Sent: Tuesday, June 25, 2013 9:20 AM
>Subject: Re: z/OS subroutine in assembler, used in both batch & CICS , making
>re-entrant
>
>
>On Tue, Jun 25, 2013 at 9:56 AM, Tom Marchant <[email protected]>wrote:
>
>> On Tue, 25 Jun 2013 08:15:46 -0500, John McKown <
>> [email protected]> wrote:
>>
>> >1) Can I use a BAKR instruction in CICS successfully.
>>
>> You will still need to provide a save area, which would be marked with
>> "F1SA"
>> in offset 4, or set register 13 to zero. This is necessary for debugging
>> purposes,
>> even if you never call any other program that will require a save area.
>> See
>> Chapter 2 of the Assembler Services Guide for release 12 or above of z/OS.
>>
>> >4) What about not using a save area at all?
>>
>> Please don't do that.
>>
>
>Looks like I'm going to be forced to. That is, use the caller's save area
>and not set up one of my own. I don't use any z/OS services (SVC, PC) or
>even macros in the code. The code gets a buffer and modifies it using only
>CPU instructions. Otherwise, I must either abandon the attempt altogether;
>which makes debugging harder, but not impossible; or make the HLASM
>subroutine LE compliant. I don't want to do this. It might be considered
>"disruptive" and it is definitely "new and unproven" and so close to
>anathema to those above me and those who must use it. This is one reason to
>recode into COBOL. That is "understood" and "accepted" by all. And, as a
>big plus, will get me out of the support loop.
>
>Despite what has been posted about CICS not supporting doing a COBOL CALL
>to a composite linked subroutine, we do it at present for these routines.
>The only "gotcha" is that the CALL'd code cannot do any CICS requests. The
>code in question is "pure CPU" code without any subroutine calls, PCs, SVCs
>or even macro invocations.
>
>--
>
>> Tom Marchant
>>
>> ----------------------------------------------------------------------
>> For IBM-MAIN subscribe / signoff / archive access instructions,
>> send email to [email protected] with the message: INFO IBM-MAIN
>>
>
>
>
>--
>This is a test of the Emergency Broadcast System. If this had been an
>actual emergency, do you really think we'd stick around to tell you?
>
>Maranatha! <><
>John McKown
>
>----------------------------------------------------------------------
>For IBM-MAIN subscribe / signoff / archive access instructions,
>send email to [email protected] with the message: INFO IBM-MAIN
>
>
>
----------------------------------------------------------------------
For IBM-MAIN subscribe / signoff / archive access instructions,
send email to [email protected] with the message: INFO IBM-MAIN
