Re: vms argv[0] and exit status fixes.

[John E. Malmberg]

New patch attached as described below.

On 3/5/2014 3:30 PM, h.becker wrote:
> I'm still looking at this, so I may change my mind or suggest to change
> the code :-)
>
> The main wrapper:
>
> There were/are problems with the program name, however I'm not sure
> whether the wrapper solves them. For DCL the wrapper, as before,
> requires a foreign command pointing to the executable (or the logical
> DCL$PATH pointing to the directory with the matching executable).
<snip>
> but when you want to use a different symbol:
>
> $ del/symb make
> $ gmake := $USR_ODS5:[BECKER_H.make.make]make.exe
> $ gmake -f MAKE_COMMAND.mf
>   make
> $ gmake -f MAKE_COMMAND.mf x
> make -f x.mf
> %DCL-W-IVVERB, unrecognized command verb - check validity and spelling
>   \MAKE\
> $

There is no way that I know of to know what foreign symbol was used to 
invoke an image.  All we get is the image path in one of three formats 
based on what the decc$ feature settings.

> I already made some changes for this, but I'm not yet convinced - it's
> not really nice and there may be other side effects:

I think that a solution is to convert the original arvg[0] to VMS format 
if needed and then use lib$set_symbol() to create a foreign command that 
is visible to the child programs with the new argv[0].

So lets try that with the make symbol missing, but a gmake symbol defined.

EAGLE> make -f makecommand.mf x
%DCL-W-IVVERB, unrecognized command verb - check validity and spelling
 \MAKE\
EAGLE> gmake -f makecommand.mf x
make -f x.mf
 here we are

The side effect is that it leaves a local symbol make behind.  Not too bad.

That source module is or can be a common source file for the latest 
ports of GNU bash, grep, and sed.

I made setting the symbol a compile time option, since those other 
programs do not need it set.

> SYS$SCRATCH:
>
> I know that the VMS CRTL does some special handling for /tmp. As far as
> I know, this is translated to SYS$SCRATCH:. So there seems to be no need
> to replace sys$scratch with /tmp. But tmp (or TMP?) may be a logical
> name pointing to a different directory and then that directory is used
> as scratch directory. Whether that is a good thing or just confusing, is
> another question. SYS$SCRATCH seems more obvious to me.

If someone has defined /TMP to other than SYS$SCRATCH: then they will 
probably be suprised if a GNU utility is not honoring it and still using 
SYS$SCRATCH:.

> exit -> vms_exit:
>
> Looks good to me, except MAKE_FAILURE is mapped twice and MAKE_TROUBLE
> isn't mapped.

A bug, easily fixed.

> If possible I would move the vms_exit function into a vms specific C
> source, either an existing one or a new one. I would move any other vms
> wrapper in such a module as well and reduce the to be included stuff to
> "usual" header file content. (OK, I know there is vmsjobs.c)

If we move it out of a header file, it is no longer an in-lined static 
function.

If die() is the only thing that calls exit, we can put it in the main.c 
module.

The other place for it would be in the config.h-vms.template file.

With the previous patch posted:

270 Tests in 75 Categories Failed (See .diff* files in work dir for 
details) :-(

With the current patch posted:

270 Tests in 75 Categories Failed (See .diff* files in work dir for 
details) :-(

No changes.  I think I finally have the test harness running correctly 
on VMS.

Regards,
-John

0001-VMS-argv0-and-exit-status-fix-revision-1.patch.gz
Description: application/gzip