开云体育

Anyone had any success in getting any of these compilers to work? I am trying Fortran G (FO520). The source for IEYUNF is cut short, and there is no MACRO library so the provided source won't Assemble.

On 9/2/20 4:54 AM, naturtle wrote:

Anyone had any success in getting any of these compilers to work? I am trying Fortran G (FO520). The source for IEYUNF is cut short, and there is no MACRO library so the provided source won't Assemble.

The damaged IEYUNF has been replaced on the version of the OS/360 CD-ROM image available from my site:

Phil Roberts and I worked on getting the FORTRAN compilers to assemble several years ago and found the undamaged modules for FORTRAN in the MTS distribution.

I was able to use all of them from the object libraries in the DLIB set, but only was able to get the COBOL compilers (both E and F) to assemble and work in a z/OS 2.2 system.? Interestingly, the LE library modules worked as direct replacements for the COBOL F library modules also, but I only found that interesting, not necessarily useful.

I tried to assemble the FORTRAN and PL/I compilers, but was never successful in getting a running program.

Dennis

--
Dennis

Is there JCL somewhere in that ISO to build the FORTRAN compiler?

Bill

On 9/3/20 9:36 AM, Bill Lewis wrote:

Is there JCL somewhere in that ISO to build the FORTRAN compiler?
Bill

Bill,

Not in the ISO, no, but you can look at this page to see the progress I made on the FORTRAN G and FORTRAN H compilers:



Beneath the section for each there is a link to an archive containing a DSSDUMP tape which also includes the jobstream to restore the dumped datasets. In order to restore the datasets it will likely be necessary to modify the jobstream to fit your system.

I have not revisited any of this since 2016 when I dumped the datasets and uploaded them to this tape. As I stated on that page, I was able to build the compilers from the source and they seem to perform as expected when compiling test programs. But I believe that they are not an exact match for the compilers distributed with the MVT system (in object code format only) and I did not continue on with any efforts to work on that.

Certainly in order to make any progress on building the FORTRAN H from the source you will need to use the undocumented 'magic' of the XL parm.

Regards,
Jay

On Thu, Sep 3, 2020 at 11:15 AM, Jay Moseley wrote:

On 9/3/20 9:36 AM, Bill Lewis wrote:

Is there JCL somewhere in that ISO to build the FORTRAN compiler?

Bill

Bill,

Not in the ISO, no, but you can look at this page to see the progress I
made on the FORTRAN G and FORTRAN H compilers:



Beneath the section for each there is a link to an archive containing a
DSSDUMP tape which also includes the jobstream to restore the dumped
datasets. In order to restore the datasets it will likely be necessary
to modify the jobstream to fit your system.

I have not revisited any of this since 2016 when I dumped the datasets
and uploaded them to this tape. As I stated on that page, I was able to
build the compilers from the source and they seem to perform as expected
when compiling test programs. But I believe that they are not an exact
match for the compilers distributed with the MVT system (in object code
format only) and I did not continue on with any efforts to work on that.

Certainly in order to make any progress on building the FORTRAN H from
the source you will need to use the undocumented 'magic' of the XL parm.

Regards,
Jay

Cool, thanks!

Bill

Wow - not only have you been down this rabbit hole, but have done a huge amount of work. I have found your MVTSRC.3350 which contains the FORTG.MACLIB. Fantastic. Amazing. I hope these materials will answer all my questions. Thanks again.