Re: COBOL COMPILER BLUES... - Hewlett Packard

This is a discussion on Re: COBOL COMPILER BLUES... - Hewlett Packard ; Yes guys, I got that - thanks. On the other hand, I think the reason I have to build that file first is because there is an error in the compiler or the compiler script, not for any intrinsic reason ...

+ Reply to Thread
Results 1 to 2 of 2

Thread: Re: COBOL COMPILER BLUES...

  1. Re: COBOL COMPILER BLUES...


    Yes guys, I got that - thanks.

    On the other hand, I think the reason I have to build that file first is because there is an error in the compiler or the compiler script, not for any intrinsic reason related to file processing or the way it outputs to standard list. There are workarounds, and I seriously doubt anyone is maintaining the COBOL compiler, and it isn't that much of an inconvenience.

    The reason, by the way, I habitually keep list files is that they come in quite handy in cases where I have to investigate running code, usually with some sort of debugger, and usually at a remote site. The CODE, MAP, and CROSSREF options really can come in handy.

    And they come in handy matching compile time and date to the object files on a system, when you have to prove what you have during a SOX audit.

    Also, using the CODE option, if one can get it to work, is a great way to learn the assembler language for a platform.
    $CONTROL CODE doesn't seem to generate an assembler listing for some reason though. Have to go read the manual.

    None of which really applies here, so I suppose I could eliminate the overhead, but I am comfortable with it, so I will probably just keep it. As un-HP3K as it may be.


    -Paul



    On Wednesday, October 29, 2008, at 04:42PM, "Matthew Perdue" wrote:
    >Quoting Duane Percox (in part):
    >
    >"If Paul wants to have his output listing be a file he has to prebuild
    >it..."
    >
    >Yes, I said that in my 10-29-2008 1:33AM post:
    >
    >"Guys...
    >
    >You're getting the error because in the first way used, the Cobol compiler
    >expects the list file to already exist (Paul's first example):
    >
    > :COB85XL DISCTST,DISCTST.OBJ.RSEN,DISCTST.LST.RSEN
    >
    >tells the compiler to direct output to disctst.lst.rsen and the compiler
    >expects the file to already exist in the permanent domain. Since the file
    >doesn't, the error "NONEXISTENT PERMANENT FILE (FSERR 52)" is reported and
    >the compiler aborts."
    >
    >* To join/leave the list, search archives, change list settings, *
    >* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *
    >
    >


    * To join/leave the list, search archives, change list settings, *
    * etc., please visit http://raven.utc.edu/archives/hp3000-l.html *


  2. Re: COBOL COMPILER BLUES... CODE option

    > Also, using the CODE option, if one can get it to work, is a great way to
    > learn the assembler language for a platform.
    > $CONTROL CODE doesn't seem to generate an assembler listing for some
    > reason though. Have to go read the manual.


    FWIW:
    As I've seen, the CODE option for old Classic/Cisc compilers, did output the
    assembly to the listing file.
    I've seen spl, fortran, pascal, and cobol listings - as I recall.

    But under Native/PARisc, I believe only the Pascal compiler has the option
    to output PArisc assembly.
    For I have not seen a working option on the other compilers to output PARisc
    code.

    But I know there are undocumented switches/options that exist for the other
    compilers.

    Keven

    * To join/leave the list, search archives, change list settings, *
    * etc., please visit http://raven.utc.edu/archives/hp3000-l.html *


+ Reply to Thread