Re: Find control boundary with a machine interface instruction

Thanks anyway, Marco

I think I already have a proof of concept ready.
Javier Mora

El mié, 12 feb 2025 a las 18:16, Marco Facchinetti (<
marco.facchinetti@xxxxxxxxx>) escribió:

Hi Javier, sorry but I'm not a MI friend :-)

When you asked about ctl boundaries I was looking how to test program state
(we have a problem with sockets pgm's) that's why I was so prompt :-)

Best regards
--
Marco Facchinetti

Mr S.r.l.

Tel. 035 962885
Cel. 393 9620498

Skype: facchinettimarco


Il giorno mer 12 feb 2025 alle ore 12:27 datil400 <datil400@xxxxxxxxx> ha
scritto:

Hi Marco,

I have not tested the performance of both options in a controlled
environment. Juenli Liu says that the MI solution is much faster than the
API. Keep in mind that the latter is always a dynamic call to program,

with

the cost that implies.

Another issue, do you have experience in the use of MATINVAT? Do you know
any examples of complex use?
Best regards,

Javier Mora

El mar, 11 feb 2025 a las 19:08, Marco Facchinetti (<
marco.facchinetti@xxxxxxxxx>) escribió:

Hi Javier, just a question about QWVRCSTK.

Since we use it a lot in trigger programs to avoid actions (recursive
calls) if the trigger invocation is caused by specific programs I'd

like

to

know how expensive can be the API instead of MI.

TIA
--
Marco Facchinetti

Mr S.r.l.

Tel. 035 962885
Cel. 393 9620498

Skype: facchinettimarco


Il giorno mar 11 feb 2025 alle ore 18:58 datil400 <datil400@xxxxxxxxx>

ha

scritto:

Thanks Marco,

I have read this documentation dozens of times and never paid

attention

to

this attribute.

I am going to try
Javier Mora

El mar, 11 feb 2025 a las 18:36, Marco Facchinetti (<
marco.facchinetti@xxxxxxxxx>) escribió:

Hi Javier, about MATINVAT:

19
Invocation status of the specified invocation (including invocation

flags).

Bit 0
Cancelled
Bit 1
Ending -- a return operation has been initiated from within the

invocation

or the actual termination of a cancelled invocation has begun.
Bit 2
Invocation interrupted by exception
Bit 3
Invocation interrupted by event (reserved)
Bit 4
Invocation is a non-bound program CALLX exception handler
Bit 5
Invocation contains a non-bound program CALLI exception handler
Bit 6
Invocation contains a signalled non-bound program branchpoint

handler

Bit 7
Retry not allowed
Bit 8
Resume not allowed
Bit 9
Resume point has been modified
Bit 10
Invocation is a program entry procedure and is marked as the oldest

in

the

activation group (This is also known as a hard control boundary.)
Bit 11
Invocation is a soft control boundary.
Bit 12
Invocation created an unnamed activation group.
Bits 13-15
Reserved
Bits 16-31
Invocation flags
(4 bytes.)
Performance consideration: When the only invocation status

information

required is the invocation flags, there may be a significant

performance

advantage if the following attribute is materialized instead of

this

one.

HTH
--
Marco Facchinetti

Mr S.r.l.

Tel. 035 962885
Cel. 393 9620498

Skype: facchinettimarco


Il giorno mar 11 feb 2025 alle ore 16:08 datil400 <

datil400@xxxxxxxxx>

ha

scritto:

Hi Martijn,

thank you for answering. I already have a utility that allows me

to

load

and traverse the call stack using the QWVRCSTK API. It is valid

but

has

low performance.


I am looking for a faster solution where I don't need to load the

entire

call stack. It is a little challenge I have set myself.

Best regards

Javier Mora

El mar, 11 feb 2025 a las 13:49, Martijn van Breden (<
m.vanbreden@xxxxxxxxxxxxxxxxxxxxxxxxxx>) escribió:

On the risk of being accused that whatever the question is, SQL

is

the

answer... 😉

Have you tried the table function Stack_info(). It does present

you

control boundaries. It may not be as fast as MI api's but

chances

are

that

your successors love you for using SQL instead of api's 🙂


Kind regards,



Martijn van Breden

lead software architect





Hi Jon,

In 2010, Junlei Li published an article explaining how to

traverse

the

call

stack using MI. Very fast and efficient, no need to load the

entire

call

stack. That utility provided a lot of information about the

call

stack

entry, but not the ‘control boundary’.

I have been using a self-developed tool for exception handling

in

RPG

for

several years, in which I use the Junlei utility. Now I need to

find

out

the 'control boundaries', at least the nearest one, for some

procedures

responsible for sending messages.

My idea is to simulate the new %target(*CTLBDY...) built-in in

my

exception

handling tool.

As a challenge I find it very interesting. If there is no other

choice, I

will use the QWVRCSTK API.
Best regards,

Javier Mora

El lun, 10 feb 2025 a las 18:46, Jon Paris (<

jon.paris@xxxxxxxxxxxxxx

)

escribió:

Perhaps if you told us _why_ you want this and what you plan

on

using

it

for we might be able to offer some alternative suggestions.


Jon Paris
Jon.Paris@xxxxxxxxxxxxxx

On Feb 10, 2025, at 12:44 PM, datil400 <datil400@xxxxxxxxx

wrote:

Thanks Marco for your help.


With the instructions FNDRINVN, MATINVAT and MATPTRIF, I am

trying

to

get

whether any stack entry is a control boundary or not. I am

not

finding

this

information when materializing the corresponding pointers.

I think I could deduce it from the activation group mark by

comparing

it

with the previous entry.

I don't know if I'm on the right track.
Javier Mora

El lun, 10 feb 2025 a las 17:38, Marco Facchinetti (<
marco.facchinetti@xxxxxxxxx>) escribió:

https://www.ibm.com/docs/en/i/7.5?topic=execution-materialize-invocation-stack-matinvs

But don't forget:

Warning
The following information is subject to change from

release

to

release.

Use

it with caution and be prepared to adjust for changes with

each

new

release.

Note
It is recommended that you use Materialize Invocation

Attributes

(MATINVAT)

or Materialize Invocation Entry (MATINVE) to materialize

an

8-byte

invocation mark. 4-byte marks can wrap and produce

unexpected

results.

--
Marco Facchinetti

Mr S.r.l.

Tel. 035 962885
Cel. 393 9620498

Skype: facchinettimarco


Il giorno lun 10 feb 2025 alle ore 17:06 datil400 <

datil400@xxxxxxxxx

ha

scritto:

Hi guys,

QWVRCSTK API and STACK_INFO table function return

information

about

the

control boundary of each entry in the call stack.

I have been investigating to find out if it is possible

to

obtain

this

data

through the machine interface. I have not found anything

that

helps

me.

Is this possible?
Best regards

--
Javier Mora
https://github.com/datil400
--
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.

--
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.

--
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.

--
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.

--
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.

--
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.

--
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.

--
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.

--
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.

--
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.

--
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.

--
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.