Those are called heisenbugs.

Interesting.

This points to the original problem being a memory corruption issue. When
problems "disappear" when debugging it is because the initiation of the
debug process has caused a change in the memory map. The result is that the
memory corruption that caused the original error is now corrupting
something else - but it just doesn't cause an explosion.

Check the length of the variables being used in all cases where you are
passing the adress as a pointer and/or are manually supplying a length.
Chances are that you are telling some routine or other to copy (say) 1,000
bytes when the variable in question is only 500 long.

In fact, just check the length of everything you are passing and make sure
there is no possibility of overflow.


Jon P.

On Jul 3, 2024, at 6:26 PM, James H. H. Lampert via MIDRANGE-L <

midrange-l@xxxxxxxxxxxxxxxxxx> wrote:

On 7/3/24 10:55 AM, Mark Waterbury wrote:

See:

https://www.ibm.com/support/pages/system/files/inline-files/i-mysterious_application-pdf.pdf

Not being a big user of environment variables on Midrange boxes (I have

only the vaguest memory of any previous Midrange usage), I put off
following Mark's link, and then, once I'd skimmed over it enough to realize
how to use the

"Debug memory manager," I put off acting on it a bit longer.

Eventually, I found WRKENVVAR, and created a QIBM_MALLOC_TYPE envvar,

set to 'DEBUG.'

The malfunction didn't happen. At all.

I then tried again, setting it to 'QUICKPOOL.' Again, no malfunction.

I then tried again, setting it to 'DEFAULT.' Still no malfunction.

I even set it to 'FOOBAR,' and then deleted it entirely. Still no

malfunction.

It seems that as long as QIBM_MALLOC_TYPE *exists* for a job, and *has

at some time had a non-null value* in that job, the malfunction does not
occur. But if, within the context of a given job, it has never existed with
a non-null value, then the malfunction occurs.

This is weird. Particularly given that the overwhelming majority of my

HTTPAPI development takes place on this box, and it's never happened before
this project. Does any of this make any sense to anybody else here?

--
JHHL
--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing

list

To post a message email: MIDRANGE-L@xxxxxxxxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
visit: https://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request@xxxxxxxxxxxxxxxxxx
Before posting, please take a moment to review the archives
at https://archive.midrange.com/midrange-l.

Please contact support@xxxxxxxxxxxxxxxxxxxx for any subscription

related questions.

--
This is the Midrange Systems Technical Discussion (MIDRANGE-L) mailing list
To post a message email: MIDRANGE-L@xxxxxxxxxxxxxxxxxx
To subscribe, unsubscribe, or change list options,
visit: https://lists.midrange.com/mailman/listinfo/midrange-l
or email: MIDRANGE-L-request@xxxxxxxxxxxxxxxxxx
Before posting, please take a moment to review the archives
at https://archive.midrange.com/midrange-l.

Please contact support@xxxxxxxxxxxxxxxxxxxx for any subscription related
questions.