Are you sure you are running compiled in both instances? Compiled Rexx will 
silently fall back to interpreted if the right library is not allocated.

Charles


-----Original Message-----
From: IBM Mainframe Discussion List [mailto:[email protected]] On Behalf 
Of David Crayford
Sent: Wednesday, September 12, 2018 3:25 AM
To: [email protected]
Subject: Re: Compiled REXX performance

This is totally bizarre! I'm getting different results running in TSO 
foreground and IKJEFT01 batch running the same compiled REXX program.

TSO

CPU time = 2.19

Batch

CPU time = 0.05

                                         CPU (Total)  Elapsed CPU 
(TCB)    CPU (SRB)     Service
Jobname  Stepname ProcStep    RC    I/O hh:mm:ss.th  hh:mm:ss.th 
hh:mm:ss.th  hh:mm:ss.th     Units
BIGREXC  BIGREXC              00     81       00.10 00.18        
00.10        00.00      1340


On 12/09/2018 5:55 PM, Mike Shorkend wrote:
> I use the procedure provided by IBM, taking all compiler option defaults
>
> // JCLLIB ORDER='FAN140.SVSC.PROCLIB'
> //S1 EXEC REXXC
> //SYSCEXEC DD DSN=MIKE.CREXX(LOOP2),DISP=SHR
> //SYSIN    DD DSN=MIKE.REXX(LOOP2),DISP=SHR
>
> which expands to
>
>
> XX*-----------------------------------------------------------------
>
>        XX* Compile REXX
> program.
>
> XX*-----------------------------------------------------------------
>
>
> XX*
>
>      5 XXREXX    EXEC
> PGM=REXXCOMP,PARM='&OPTIONS'
>        IEFC653I SUBSTITUTION JCL - PGM=REXXCOMP,PARM='XREF
> OBJECT'
>      6 XXSTEPLIB   DD
> DSN=&COMPDSN,DISP=SHR
>        IEFC653I SUBSTITUTION JCL -
> DSN=FAN140.SFANLMD,DISP=SHR
>      7 XXSYSPRINT  DD
> SYSOUT=*
>      8 XXSYSTERM   DD
> SYSOUT=*
>        XX*SYSIEXEC DD
> DUMMY
>        XX*SYSDUMP  DD
> DUMMY
>      9 //SYSCEXEC DD
> DSN=MIKE.CREXX(LOOP2),DISP=SHR
>        X/SYSCEXEC  DD
> DSN=&&CEXEC(GO),DISP=(MOD,PASS),UNIT=SYSDA,
>        X/
> SPACE=(800,(800,100,1))
>     10 XXSYSPUNCH  DD
> DSN=&&OBJECT,DISP=(MOD,PASS),UNIT=SYSDA,
>        XX
> SPACE=(800,(800,100))
>     11 //SYSIN    DD
> DSN=MIKE.REXX(LOOP2),DISP=SHR
>
>
>
>
>
>
> On Wed, 12 Sep 2018 at 12:44, David Crayford <[email protected]> wrote:
>
>> z/OS 2.3, IBM Compiler for REXX on zSeries 4.0  LVL PI76785.
>>
>> Can you share you compiler JCL?
>>
>>
>> On 12/09/2018 5:32 PM, Mike Shorkend wrote:
>>> David - I ran it on z/OS 2.2, the REXX compiler is 1.4
>>> How about you?
>>>
>>> On Wed, 12 Sep 2018 at 12:27, David Crayford <[email protected]>
>> wrote:
>>>> On 12/09/2018 4:56 PM, Mike Shorkend wrote:
>>>>> The new test shows:
>>>>> Non-compiled
>>>>>
>>>>> CPU TIME = 77.09
>>>>>
>>>>> Compiled
>>>>>
>>>>> CPU TIME = 11.27
>>>> Your results are different to mine!
>>>>
>>>> Interpreted:  CPU time = 3.12
>>>>
>>>> Compiled:     CPU time = 7.99
>>>>
>>>>> Not so great, because of the inefficient memory management with stem
>>>>> usage?
>>>> ----------------------------------------------------------------------
>>>> 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
>>
>

----------------------------------------------------------------------
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

Reply via email to