From: bleau@umtof.umd.edu
Sent: Wednesday, July 12, 2000 2:58 PM
To: Info-VAX@Mvb.Saic.Com
Subject: Re: MX 4.2 won't rebuild, what's missing?

In article <396be73e.524144494f47414741@radiogaga.harz.de>, martin@radiogaga.harz.de (Martin Vorlaender) writes:
>:
>: Sure enough, there isn't a STARLET.B32 (or whatever it's supposed to be
>: called) in SYS$LIBRARY:.
>
>$ BLISS /LIBRARY=SYS$COMMON:[SYSLIB] SYS$LIBRARY:STARLET.REQ
>
>(This is done during BLISS installation on a VAX system.)

If you note my original post, I state that I'm on an Alpha, not a VAX.

Another poster, who replied on vmsnet.mail.mx, gave me the magic formula to
create the above library:

>     $ set def sys$common:[syslib]
>     $ bliss/library starlet.req
>     $ bliss/librar=lib.l32 starlet.req+lib.req

This works, just ignore the information messages you get.

I've had other problems since then, though, in trying to rebuild MX 4.2 on
Alpha.  I noticed the admonition in ther readme file to build FLQ before
COMMON, so I started that way.  Doing a MMS in the [MX.FLQ] directory gave me:

BLISS /OBJ=FLQ_OPEN.OBJ/TERM=STAT/NOLIST/NODEBUG/VAR="" FLQ_OPEN.B32
%DCL-W-IVKEYW, unrecognized keyword - check validity and spelling
 \STAT\
%MMS-F-ABORT, For target FLQ_OPEN.OBJ, CLI returned abort status: %X00038060.
-CLI-W-IVKEYW, unrecognized keyword - check validity and spelling

I assume this is a minor problem and just removed the /TERM=STAT from the
BFLAGS lines, then invoked MMS again.  This time, though, I got a slew of
errors on the module FLQ_VECTOR.MAR:

MACRO /OBJ=FLQ_VECTOR.OBJ/NODEBUG FLQ_VECTOR.MAR

        ROUTINE FLQ_OPEN
^
%AMAC-W-DIRTRANS, .TRANSFER directive is not supported
-MACRO-I-ATEXPLINE, at macro expansion line 3
at line number 29 in file MX_SRC:[MX.FLQ]FLQ_VECTOR.MAR;1
        .TRANSFER       FLQ_OPEN

        ROUTINE FLQ_OPEN
^
%AMAC-E-DIRMASK, .MASK directive is not supported
-MACRO-I-ATEXPLINE, at macro expansion line 4
at line number 29 in file MX_SRC:[MX.FLQ]FLQ_VECTOR.MAR;1
        .MASK           FLQ_OPEN

        ROUTINE FLQ_OPEN
^
%AMAC-E-BLKEXPNABS, block expression not absolute
-MACRO-I-ATEXPLINE, at macro expansion line 4
at line number 29 in file MX_SRC:[MX.FLQ]FLQ_VECTOR.MAR;1

[snip]
        .END
^
%AMAC-E-PREVERROR, previous errors prevent further analysis
at line number 42 in file MX_SRC:[MX.FLQ]FLQ_VECTOR.MAR;1


I tried the above using the /MIGRATION qualifier and got the same results.

I checked the help file and found this:

MACRO

  /MIGRATION

    Support_of_VAX_MACRO_Assembler_Directives

         The compiler supports most of the standard VAX MACRO assembler
         directives. However, the following directives that are supported
         by the VAX MACRO assembler do not make sense for compiled code.
         Consequently, the compiler flags them and continues execution.
         You can disable the flagging of these directives by specifying
         /NOFLAG=DIRECTIVES.
[snip]
         o  .TRANSFER

         o  .MASK
[snip]

So it seems the directives really aren't needed for Alphas.  Does this mean I
can change the FLQ's DESCRIP.MMS file to use /MIGRATION/NOFLAG=DIRECTIVES
and just rerun MMS?  Or are the above messages really errors?

Meanwhile:
>BLISS /OBJ=FLQ_OPEN.OBJ/TERM=STAT/NOLIST/NODEBUG/VAR="" FLQ_OPEN.B32
>%DCL-W-IVKEYW, unrecognized keyword - check validity and spelling
> \STAT\
>%MMS-F-ABORT, For target FLQ_OPEN.OBJ, CLI returned abort status: %X00038060.
>-CLI-W-IVKEYW, unrecognized keyword - check validity and spelling

I think I found the reason for this.  There's a macro __ALPHA__ that is tested
in the DESCRIP.MMS file for FLQ (and probably elsewhere as well).  The readme
file doesn't say anything about setting/defining this macro.  When I defined
it, though, using /MACRO=__ALPHA__=1, it didn't give me any errors.  As a
matter of fact, it didn't compile FLQ_VECTOR.MAR at all!  I used the following
command

MMS/MACRO=__ALPHA__=1/NOACTION/OUT=X./FROM/DESCRIP=DESCRIP.MMS;1

to ensure the original makefile was used, to force all compilations to be done
again (in case some were compiled incorrectly), and to get a list of the
commands that would be done.  The MACRO command doesn't even appear now!

The problem I now have, though, is that the module FLQ_VECTOR.OBJ is listed in
the options file MX_FLQ_SHR.OPT which is specified on the LINK command.  Since
I cleaned up the directory (ie, deleted old .OBJ files), and MACRO is no longer
invoked, LINK now fails due to a missing file. <sigh>

I'll try compiling FLQ_VECTOR.MAR with the qualifiers I mentioned in my last
post.  Meanwhile, what *is* the correct answer?

Lawrence Bleau
University of Maryland
Physics Dept., Space Physics Group
301-405-6223
bleau@umtof.umd.edu