Accessing Main Storage: CORE › Display Main Storage › CORE Commands For PL/I Symbolic Programs

CORE Commands For PL/I Symbolic Programs

If you have purchased the PL/I Symbolic Option and are currently stopped at a breakpoint, use CORE to view storage associated with a symbolic PL/I program. The PL/I Symbolic CORE facilities work just like the COBOL Symbolic facilities of CORE, with the following differences:

• There are three PL/I symbolic formats for CORE commands:

  CORE='symbolname'
  

  This command checks to see if the symbol is a label. If it is, the machine code at that location is displayed. If it is not, the definition of the symbol in the most recently activated procedure or block is used to display the current value of the symbol.

  Note: The most recently activated block, containing a definition of the symbol, need not be the currently executing block.

  CORE='blockname:symbolname'
  

  CORE='blockname:name1.name2.name3'
  

  This command checks to see if the symbol is a label in the qualifying block. If the named block is active, the definition of the symbol in that block is used to display the current value of the symbol. If it is not, the blockname is checked to verify that it is on the active block chain.

  CORE='symbolname1 >symbolname2'
  

  This command displays variables listed as based (*) in the PL/I Cross Reference, where:

  • symbolname1 is assumed to be a pointer that points to an area to be mapped by symbolname2. If it is not a pointer, then the address of symbolname1 is mapped by symbolname2.
  • symbolname2 formats the data to be displayed. In the case of a structure, symbolname2 is assumed to be the beginning of the area pointed to by symbolname1.

    CORE='name1.name2(name3,name4)'
    

  Display array item indicated by the resolution of variables name1 and name2. Up to 15 levels of qualification and subscripting are supported.

• Controlled variables (CTL) are not supported.
• Symbolic CORE commands should not be used when a program is stopped at a PROC or BEGIN statement. The DSA environment is not properly established by PL/I until the first executable statement in a block is actually executed. As a result, CORE provides unpredictable results if you refer to a symbol that has not been allocated or freed (such as BASED variables that are allocated in the block).
• Like named variables are supported as follows:
  • If two variables with the same name are declared in the same block/procedure, attempting to display these variables with a CORE command requires qualification.

    CORE='name1.name2'
    

  • If two variables with the same name are declared in different blocks/procedures, display them using the format:

    CORE='blockname:symbolname'
    

• Based variables are supported with an unlimited number of levels of basing.

Example

See this sample in the following examples.

PROC1:  ...;

     DCL 1    A,
              2 FIELDB  ...,
                3 FIELDC ...;
   SUBPROC:  ...;
     DCL 1   A,
             2 FIELDB  ...,
               3 FIELDC ...;
   END SUBPROC;

To display the variable FIELDB declared in PROC1, enter:

CORE='PROC1:FIELDB'

To display the variable FIELDB declared in SUBPROC, enter:

CORE='SUBPROC:FIELDB'

See the following sample for the next example.

DCL   Q PTR;
DCL   R PTR;
DCL   Z BASED(Q);

To display Z using R as the pointer, code the CORE command as follows:

CORE='R >Z'