Date: Tue, 18 Jul 2000 11:57:11 +0300 (IDT)
From: Eli Zaretskii <eliz AT is DOT elta DOT co DOT il>
X-Sender: eliz AT is
To: Laszlo Molnar <laszlo DOT molnar AT eth DOT ericsson DOT se>
cc: djgpp-workers AT delorie DOT com
Subject: Re: DJGPP problem executing a script
In-Reply-To: <20000718102328.U772@libra.eth.ericsson.se>
Message-ID: <Pine.SUN.3.91.1000718114950.1391J-100000@is>
MIME-Version: 1.0
Content-Type: TEXT/PLAIN; charset=US-ASCII
Reply-To: djgpp-workers AT delorie DOT com
Errors-To: nobody AT delorie DOT com
X-Mailing-List: djgpp-workers AT delorie DOT com
X-Unsubscribes-To: listserv AT delorie DOT com
Precedence: bulk


On Tue, 18 Jul 2000, Laszlo Molnar wrote:

> On Mon, Jul 17, 2000 at 03:49:34PM +0300, Eli Zaretskii wrote:
> > > Ok, perl 5.6 builds nicely using djdev203 and gcc2952.
> > Did you also fix the problem with exit status of subordinate programs
> > which wasn't shifted 8 bits to the left?  (This causes one of the tests
> > from the GNU Make test suite to fail, unless the test script is tweaked.)
> 
> IIRC, I fixed a problem with return codes some times ago. But to be
> sure, could you send me a small test case for the problem you observed?

The following is an (untested) excerpt from tests/scripts/features/errors
script in GNU Make distribution.  If this snippet fails to exhibit the 
bug in the previous port of Perl, please try to run the Make test suite.

  unlink("cleanit");
  $cleanit_error = `sh -c "rm cleanit 2>&1"`;
  $delete_error_code = $? >> 8;
  #$delete_error_code = $?;

In the current port, the exit code was the usual DJGPP-style one, whereas 
Perls wants the exit code in the high 8 bits.  The commented-out line 
above is what I needed in the Make test suite to get this test to work.

> > > However I had a
> > > problem with bash 2.03: when I run the test suite it just stopped on a
> > > file - the first 7 subtests of 25 was OK, then execution is stuck in
> > > the 8th test until I pressed Ctrl-C.
> > Does this always happen with the 8th script, or only with the particular
> > script which happens to be the 8th in the original order?  In other
> > words, what happens if you remove the first test?
> 
> It happens in a particular script.

Perhaps it would be a good idea to post that script.

> when I start configure.bat I usually use the "-dEs" command line
> switch to make the configure process non-interactive.

I don't think I know what -dEs is, how does it affect the configuration 
process, and why do you need to press ENTER if it doesn't work.  Could 
you please explain?

Did -dEs work correctly with Bash v2.03?