Re: CWX9001 and MCH4227 when compiling FOR-EACH ... IN %SPLIT(...) -- RPG400-L

Display Formatted Message Text
Message ID . . . . . . . . . : CWX9001   Message file . . . . . . . . : QCWXMSG    Library . . . . . . . . . : QSYS        Message . . . . : An error occurred during translation.    Cause . . . . . : An irrecoverable error has occurred during translation.     The reason code is &1. See the previous messages listed in the joblog.    Recovery . . . : This error indicates that an internal compiler error has    occurred. Contact your service representative.        Display Formatted Message Text
Message ID . . . . . . . . . : MCH4227   Message file . . . . . . . . : QCPFMSG    Library . . . . . . . . . : QSYS        Message . . . . : Instruction stream not valid.    Cause . . . . . : The object was not created because the sequence of    instruction stream objects is not valid.   Recovery . . . : There is a problem in the compiler. Report this problem    to the supplier of the compiler.   Technical description . . . . . . . . : The reason code is &1. The current    instruction count is &2. The offset to the instruction stream data that is    not valid is &3 bytes. The offset to the current instruction is &4 bytes     and the operation code of the current instruction is &5. The current     statement number is &6. The LIC log note ID is &7.
Notice that CWX9001 and MCH4227 messages say right there in the second-level message text that it is an internal compiler error and to contact IBM Support.
I do not understand why people are obsessed with fooling around with the code, trying to "work around" what is obviously a defect.
Just saying.

On Tuesday, December 10, 2024 at 03:14:10 PM EST, Charles Wilt <charles.wilt@xxxxxxxxx> wrote:

Sounds like a bad PTF to me..

I'd contact IBM support.

Charles

On Tue, Dec 10, 2024 at 8:52 AM <daniel@xxxxxxxx> wrote:

Hi,

I have a strange problem. I have several programs in production, which use
FOR-EACH with %SPLIT. These programs are working since several months.

Now when I try to compile these programs the compiler terminates with
CWX9001 and in the job log a MCH4227 message can be found.

This is the construct that worked at least until October this year, and is
failing now:

**free
ctl-opt actgrp(*new);
dcl-s items varchar(32000) inz('1,2,3,4,5');
dcl-s item varchar(25) inz;
for-each item in %split(items:',');
dsply item;
endfor;
*inlr = *on;
return;

Now the strange things start.

If I shorten the „list“ variable to VARCHAR(16382) the program compiles:

dcl-s items char(16382) inz('1,2,3,4,5');
dcl-s item char(25) inz;
for-each item in %split(items:',');

If I declare an array and do the %SPLIT into that, the program compiles:

dcl-s items char(16382) inz('1,2,3,4,5');
dcl-s item char(25) inz;
dcl-s array char(25) dim(*auto:32000) inz;
array = %split(items:',');
for-each item in array;

What’s wrong? Any ideas?

Thanks in advance.

Kind regards,
Daniel

--
This is the RPG programming on IBM i (RPG400-L) mailing list
To post a message email: RPG400-L@xxxxxxxxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
visit: https://lists.midrange.com/mailman/listinfo/rpg400-l
or email: RPG400-L-request@xxxxxxxxxxxxxxxxxx
Before posting, please take a moment to review the archives
at https://archive.midrange.com/rpg400-l.

Please contact support@xxxxxxxxxxxxxxxxxxxx for any subscription related
questions.