Sunday, 10 September 2017

ILE

1)   Binding
2)   Modularity
3)   Reusable components
4)   Common Run time services (Service programs)
5)   Better control over resources (Activation groups)
6)   Better code optimization
7)   Foundation for the future

Binding: - Binds helps in reducing the overhead in calling programs. Binding the modules speeds up the call. To differentiate between the two types of calls, the earlier method is called as dynamic or external program call and the ILE method is called as static or bound procedure call. ILE allows you to bind modules written in different languages like RPG, COBOL, and C. 

Binding two ways
1)    Dynamic Binding – RPG/400 ( CALL pgm b from pgm A)
2)    Static Binding – ILE (Binding the modules)

Modularity: - Modular approach to application programming benefits in faster compile time, since smaller codes compile faster than a code of 2000 lines where in there was a change in 2 lines of code. If we have to update/change a large program first we have to understand exactly what it is being done. The style of different programmers varies & makes it difficult to understand. While a smaller piece of code which represents a function is far easier to grasp and modify or update. Modular approach helps in testing the functions in isolation which also covers the testing in complete ( since all the input & output paths are tested).In also helps in division of Labor. It is easy to migrate code to/from other platforms.

Module Object:-
  • Debug Data
  • PEP
  • UEP

Debug data is the data necessary for debugging a runable ILE object and is optional.
PEP Program entry procedure is the compiler generated code that is the entry point of an ILE program on a dynamic call. The system adds some code to initialize the module.
UEP user entry procedure is written by the programmer is the target of the dynamic program call. It is the procedure that gets control from PEP.

When the create program CRTPGM, is specified the ENTMOD parameter allows you to select which module containing a PEP is the programs entry point. A PEP that is associated with a module that is not selected as an entry point for the program is ignored. All other procedures and data items of the module are used as specified and only the PEP is ignored.

Program Object:-

1) Single Module programs

CRTBNDRPG or 14

2) Multiple Module programs

CRTxxxMOD (15) then CRTPGM (14)

CRTBNDRPG creates a module object in the QTEMP library and is deleted at the end of the job. CRTBNDRPG provides a single step approach to program creation.

In case of multiple module programs ILE supports 2 types of static binding. In case of Bind by copy the module containing the procedure code is copied into a new program object. When the program is created the system copies the necessary code to make the program object executable. This code is known as PEP. This is the entry point for an ILE program. In a *PGM many modules have UEPs. You specify which one is the UEP for your program object in the ENTMOD parameter.

CALL & CALLB

CALL
-          *PGM
-          Dynamic binding
-          ILE or OPM

CALLB
-          Bound Procedure
-          Static Binding
-          Module

Why Service Program??

Bind by Copy
-          Many copies of code
-          Complex Maintenance

Service Programs
-          Single Copy of code
-          Similar to subroutine or procedure library

Call performance is similar in both.

A service program is a collection of runnable procedures and available data items easily & directly accessible by other ILE programs or service programs.

Creating Service programs:-

-          Source Member
-          Modules
-          CRTSRVPGM
To create a service program, you must first create the modules that are to be copied into the service program. One or more modules from any ILE language are copied to make the service program object. Then you run the CRTSRVPGM command. The *SRVPGM object is not executable but contains a set of callable procedures. There is no PEP associated with *SRVPGM. Since no PEP a dynamic program call to a service program is not valid. A modules PEP is ignored.

HITECH Lets look at the real time scenario:-

Updation

-          CRTXXXMOD
-          UPDPGM

Service Program
-          CRTXXXMOD
-          UPDSRVPGM

To update a procedure that is bound by copy, you must recreate the module using CRTXXXMOD command. Then rebind the changed module to all *PGMs that require it using the UPDPGM command. You can also use CRTPGM to achieve the same. The difference is that the UPDPGM command requires the name of the *PGM and the name of the modules only that are to be replaced.

Like the UPDPGM command, UPDSRVPGM takes a changed module and replaces the existing module to create a new copy of the service program.


Binding Directory:-

-          CRTBNDDIR
-          ADDBNDDIRE


Objname    type                      lib
QALLOC   *SRVPGM                  *LIBL
QMATH     *SRVPGM                   QSYS
QFREE        *MODULE                 *LIBL
QHFREE    *SRVPGM                  XYZ

A binding directory contains the name of the modules & service programs that you may need when creating an ILE program or service program.
A Binding Directory is a system object that is identified to the system by the symbol *BNDDIR. The Binding directory is optional, but reduces the number of parameters you must enter when creating a *PGM or *SRVPGM. It eliminates the possibility of making error such as keying a wrong *MODULE name.

WRKBNDDIR command allows you to display and work with the list of binding directories.

IMPORT & EXPORT Procedure:-

Import
-          CALLB to procedure not bound by copy
-          Reference to procedure outside real module of *PGM or *SRVPGM

Export
-          Make procedure available to be called from outside this *SRVPGM

When you import a procedure you are calling this procedure that is not bound by copy to this *PGM object an import can also be called as reference. An import request is made when you bind by reference to a *SRVPGM

When you export a procedure in ILE, you make it available to be bound by reference to procedures outside this *PGM or *SRVPGM object. An export can also be called as the definition. A procedure is made available for export when you specify the export keyword in the prototype.

CALL OPCODES

1)    CALL – Dynamic Call
2)    CALLB – Static Bound Call
3)    CALLP – In place of CALL or CALLB, Prototyped parameters

CALL: - At the runtime, compiler is going to check for the existence of the object.

CALLB: - Compiler checks for the existence of the object at the compile time.


CALLP: - eval res = proc1(x: y: z)
Here, instead of PLIST we can send parameters directly.

You have used CALL and parameter lists to access other program objects. By using CALLB you save time because the code of the other module is statistically bound to the calling program. The parameters are defined using an interface called prototype. CALLP is one type of prototyped call.

If procedure is not returning any value we just use CALLP, but if returning any value we use expression.

PGM1          PGM2

Module1
-          P1
-          P2
  Module2
  -   P3
Module3
  -  P4
Module 4                                           
P5

 

Service Program:-

 






















                                                                                     
The best way to package sub procedures is in a service program. The steps while creating are:

1)    Code the Sub procedure source members, type RPGLE. Each sub procedure contains P spec, D spec, C spec.
2)    Use CRTRPGMOD to create a compile unit for each individual sub procedure.
3)    Use CRTSRVPGM to place all desired sub procedure modules in a *SRVPGM object.
4)    You will be able to call all sub procedures from modules bound to your service program.

Note: - A service program will have N number of modules and all modules can have N number of procedures.

Binder Language

Binder Language makes sub procedures, procedures and data names available for use by other *PGM & *SRVPGM objects by making them available for export.
Service program stores latest signature, binder language has many signatures.
So Binder language is used to store more than one signature for a particular program or module object. Ex: - Here, Binder Language is stored in QSRVSRC source file & generally the member name will be the name of the service program.

Source File QSRVSRC.Finance *PRV for old signatures

STRPGMEXP PGMLVL (*CURRENT) LVLCHK (*YES)
    EXPORT SYMBOL (TERM)
    EXPORT SYMBOL (RATE)
    EXPORT SYMBOL (AMOUNT)                                           Export Procedures
    EXPORT SYMBOL (PAYMENT)                                          &
    EXPORT SYMBOL (OPENACCT)                                        data items
    EXPORT SYMBOL (CLOSEACCT)




Signatures:-

A signature is a value that identifies the public interface supported by the service program. It is similar to level check on the AS/400 file object. You may specify your own explicit signature or the binder will generate a signature based on the list of the procedures and the data item names to be exported and the sequence in which they are specified. Therefore a signature provides a quick & convenient way to validate accessibility to the service program. The level check parameter on the STRPGMEXP command specifies during the binding process, the public interface to the service program should be checked. 
The signature is generated when the service program is created. The signature is copied to every *PGM object that binds by reference to that service program when the *PGM is created. When the bind is completed at program load, the signatures are checked, if the signature match the client referencing the service program can use the public interface with out being recompiled.
                                                                      *SRVPGM Finance
 



















OPM Model

-          RPG III
-          Basic
-          CL
-          PL/I
-          COBOL

Extended Program Model

-          C
-          PASCAL

ILE Model (Integrated Language Environment)

-          RPGIV
-          C
-          COBOL

OPM:-
Original Program Model is an integral part of OS/400. In OPM Model the program call is dynamic (CALL) that means at run time the program is going to check for the existence of the object. The variables are public across the entire program. On every compilation of a source program it creates a program object. There is only one entry point and the resources are shared across the job.

EPM:-
In Extended Program model it has multiple entry point unlike OPM. The variables are static or automatic. It is the layer above the OS/400 and thus it creates an environment which is different, ideal for development of tools.

ILE:-
ILE is also tightly integrated with OS/400 like OPM. In ILE it supports static or dynamic binding. It supports local variable and external variables. It integrates traditional language like RPG & procedure based language like C into one environment. This has resulted in a better run time behavior.

Activation Groups:-

A) Activation – What is Activation?

-          Unique Allocation
-          Changes the symbol links

After creating a program object it is ready to be executed. The process of getting a program or service program ready to run is called activation. Activation is done by the system when the program is called. But we cannot call a service program since no PEP, so they are activated during the call to a program that directly or indirectly requires their services. Each activation will be provided with one copy of static variables for each program activation. It also changes the symbolic links to service programs providing the exports into links to physical addresses.

Advantage: - No matter how many jobs are running a program or service program, only one copy of that objects instruction resides in storage. So when one program object is used concurrently by many jobs the static variables are separate for each activation. All ILE programs and service programs are activated with in an activation group.






B) Application Isolation

 




A
 
DFT AG
MENU
1. Order Entry
2. Stock Report
 
 








A
 
SR2    
 
SR1
 
           Activation A









Here, Order Entry & Stock Report are two applications.

An Activation Group is like a sub job. A job with in a job.

An Activation Group: - The method used to define Firewalls is called as Activation Group. The programmer can request that this application run in an unique activation group and that the resources (such as Data File, ODP) can be scoped to that activation group.

2) Isolates applications with in a job for achieving
-          Flexibility
-          Scoping of File overrides and commitment control
-          File sharing with in Activation Groups
3)    Selective Cleanup of applications
4)    Firewalls
5)    Applications own resources
6)    Less dependency on Job call stack






c) Program Activation – OPM

Storage for PGMA

Storage for PGMB

Storage for PGMC



 
Storage for PGMA

Storage for PGMC

 
Storage for PGMA

 
Storage for PGMA

Storage for PGMB
 
Storage for PGMA
 
                1                                                 2                                               3






              4, 5                                                         6, 7









1)    Call OPM program A, PGMA is activated.
2)    PGMA calls OPM PGMB. Now PGMB is activated.
3)    PGMB returns to PGMA with LR indicator ON. PGMB is deactivated.
4)    PGMA calls ILE PGMC, PGMC is activated.
5)    PGMC returns to PGMA with LR indicator OFF. PGMC is not deactivated.
6)    PGMA calls OPM PGMB again, PGMB is activated.
7)    PGMA calls ILE PGMC again, It now branches to PGMC (reactivation is unnecessary) since PGMA called PGMC before and there was not set on LR in PGMC at that time so it now branches to PGMC when called again from PGMA.

Note: - From this it is clear that, Activation includes the initialization of static storage. In OPM When a program ends, its activation is cleaned up.



d) Program Activation – ILE

Storage for PGMA

Storage for PGMB

Storage for PGMC

 
Storage for PGMA
Storage for PGMB

Storage for PGMC

 
Storage for PGMA

Storage for PGMB *

 
Storage for PGMA

Storage for PGMB

 
Storage for PGMA
 
              1                                                   2                                                3  

                                       



                4, 5                                         6, 7      











Note: - The above is explained with the assumption that all ILE *PGMs will be running in the same activation group.


Call ILE PGMA. The activation group is created if this the first call in the job. PGMA is activated.

PGMA calls ILE PGMB. Now PGMB is activated in the same activation group.

PGMB returns to PGMA with LR indicator ON. PGMB is not deactivated, the reinitialize flag is set.

PGMA calls PGMC. PGMC is activated in the same activation group.

PGMC returns to PGMA with LR indicator OFF. PGMC is not deactivated similar to OPM.

PGMA calls PGMB again. PGMB is not reactivated, but initialized.

PGMA calls PGMC again. PGMC is not reactivated, but initialized.

Important Note (HITEC concentrate):- ILE activation remains in the activation group in which it was initially activated for the life of that activation group, even after that program ends. If you SETON LR implicitly closed files will be closed and the static storage will be flagged for reinitialization   on the next call. But the storage is not deleted from the job.


e) Activation Groups

-          ILE programs can run in ILE activation groups.
-          OPM programs runs in default activation group
i)             ILE programs can run in the default or ILE activation groups.
ii)           Recommendation: ILE programs runs only in ILE activation groups.
-          ILE activation groups have two major purposes
i)             Protection of shared resources from other applications.
ii)           Clean up of storage by ILE programs.

All ILE programs and sub programs are activated with in a substructure of a job called an activation group. This contains the resources necessary to run the programs.


Resources fall into the following general categories:
1)    Static & dynamic program variables.
2)      Dynamic storage.
3)      Temporary data management resources.
4)      Certain type of exceptional handlers & ending procedures.

Ex: - Data File, ODP
 
 













Imp Note: - Storage for ILE programs in a job cannot be cleaned up if they are run in a default activation group. The default activation group cannot be reclaimed. Therefore, ILE activation group remains for the life of the Job. Other minor differences exist in the behavior that makes it unwise to mix OPM & ILE programs in default activation groups. RCLRSC command does not reclaim storage for ILE programs.

 f) Activation group example

For diagram see Application Isolation topic

In the above diagram, Order entry application & Stock report application run in its own activation groups OE & SR respectively. QILE is the generic activation group name that may be used for those ILE applications that do not need to run in their own activation groups.

The File A is shared resource (shared ODP). It is shared among all the programs in OE activation group & a different shared ODP is setup for all the programs in the SR activation group. The shared resource can be set across the job (all activation groups) as well as only within an activation group.

g) Activation Group creation


 



















When ever an OS/400 job is started, the system creates two default activation groups to be used by OPM programs. One activation group is reserved for system code & the other is used for all the OPM programs. You cannot delete the OPM default activation groups. They are deleted by the system when the job ends.

You can control the run time creation of an ILE activation group by specifying an activation group attributed when you create your program or service program. Based on the attribute an activation group is created by ILE as a part of dynamic program call processing. The attribute is specified by using the ACTGRP parameter on the CRTPGM command or CRTSRVPGM.

h) Activation Group attributes

Activation Groups are of three types

1)    User named activation group
2)    System named activation group
3)    Activation group of the caller

Activation group attributes

-          CRTPGM <F4> ACTGRP (*NEW)
-          CRTSRVPGM <F4> ACTGRP (*CALLER)
-          CRTBNDRPG <F4> DFTACTGRP(*YES)

User Named activation group: - A user named activation group allows you to manage a collection of ILE program and ILE service program as one application. The activation group is created when it does not already exist in the job. It is then used by all the programs & service programs that specify the same activation group name. This is specified with the ACTGRP parameter. It creates a persistent activation group.

System Named activation group: - A system named activation group allows you to create a new activation group when ever the program is called. ILE selects a name for this activation group. The name assigned by ILE is unique with in the job. The name assigned to the system named activation group does not match any name you choose for a user named activation group. ILE service programs do not support this attribute. This is specified with the ACTGRP (*NEW) parameter on the CRTPGM command. It creates a non persistent activation group.

Note: - ILE service programs do not support this attribute.

Activation group of the CALLER: - An attribute to use the activation group of the calling program allows you to create an ILE program or ILE service program that will be activated with in the activation group of the calling program. With this attribute a new activation group is never created when the program or service program is activated. This attribute is specified with ACTGRP (*CALLER) parameter.

What is meant by Persistent activation Groups? – HITEC Question

.  Programmer named activation group or user named activation group.
-          Exists until explicitly destroyed or End of Job.
-          May be destroyed because of unhandled exception.
.  Activation Group can be destroyed by
-          RCLACTGRP
-          Language specific “Hard Leave “ operations
.  COBOL: STOP Run
.  C: exit () from main
-          RPG & CL have no “Hard Leave” operations
.  CALL to ILE API CEETREC
-          Unhandled Exceptions.











No comments:

Post a Comment