From:	rupp AT gnat DOT com (Douglas B. Rupp)
Message-Id:	<9702032347.AA28728@nile.gnat.com>
Subject:	Re: bash and make gnat1 (or cc1plus)
To:	eliz AT is DOT elta DOT co DOT il (Eli Zaretskii)
Date:	Mon, 3 Feb 1997 18:47:37 -0500 (EST)
Cc:	rupp AT gnat DOT com, djgpp-workers AT delorie DOT com
In-Reply-To:	<Pine.SUN.3.91.970203164849.27099A-100000@is> from "Eli Zaretskii" at Feb 3, 97 05:06:22 pm
Mime-Version:	1.0

> 
> 
> On Sun, 2 Feb 1997, Douglas B. Rupp wrote:
> 
> > THe problem goes away if this line is removed from the command:
> > > "CC=`case 'gcc' in stage*) echo 'gcc' | sed -e 's|stage|../stage|g';; *) 
> > >    echo 'gcc';; esac`" 
> > 
> > My guess is that the expansion of this inline expression in a call to
> > system()/dosexec() is overflowing some internal buffer.
> 
> My guess is that something is wrong in the ported `bash', not in `system' 
> on the Make's side.  Did you try to see if the traceback makes sense in 
> the context of the shell?  Please post the traceback here.

I've been unable to build a version of bash.exe with symbols.  There
are numerous problems.

1) The bashb6 diffs reference nonexistent files .build and version.h.  These
   files are probably created during the build process and don't seem
   too important.

2) THe machines.h header file expects a symbol __DJGPP__, but there isn't
   such a symbol in gcc -dumpspecs.  I changed machines.h to expect GO32
   instead.

3) Make can't find the files I patched, specifically builtins/history.def.
   I don't understand this at all, I can list it with ls and dir.  THe name
   is <8 characters, so that's not even an issue.  It's not hidden, the
   date looks ok (not in the future).

The last problem has me stuck.

I think your right about this being a shell bug though.  That makes the
most sense.

- Raw text -