Re: autoconf-2.65 on UnixWare 7.1.4

[Tim Rice]

On Sat, 16 Jan 2010, Ralf Wildenhues wrote:

> Hi Tim,
> 
> * Tim Rice wrote on Wed, Jan 13, 2010 at 03:28:49AM CET:
> > On Tue, 12 Jan 2010, Ralf Wildenhues wrote:
> > > * Tim Rice wrote on Tue, Jan 12, 2010 at 01:48:43AM CET:
> > > > I have built one here and test 127 does pass with 
> > > > CONFIG_SHELL=/opt/bin/bash
> > > > so it looks like yet another ksh bug. Hopefully the UnixWare engineers
> > > > will find some time to work on my support call about the ksh bugs.
> > > 
> > > OK, so one suitable recommendation for users of this system is to
> > > install bash first.  Which bash version did you use, and does it pass
> > > all of its testsuite and does
> > >   make check TESTSUITEFLAGS=CONFIG_SHELL=/path/to/bash
> > > 
> > > pass all of Autoconf's test suite?
> >  
> > The bash I built back in 2006 was version 3.1.17(1)-release
> > With that we get testsuite: 74 211 failed
> > I started working on building bash-4.1 last night.
> > When I get that built, I'll see if it works any better.
> 
> Thanks.  I'd be interested in the global testsuite.log file, for both of
> these shells (edit out confidential information if you need to).

Attached.
bash 4.1 only failed test 211 which is interesting because if I then do
gmake -C tests clean
gmake check MAKE=gmake TESTSUITEFLAGS='-v -d -x 127 CONFIG_SHELL=/opt/bin/bash'
it passes.


[...]
> > > > iASBOX
> > > > ......
> > > > 
> > > > I've seen this before back in autoconf-2.59.
> > > 
> > > At this very same spot only, or with other here-documents or other code
> > > lines as well?
> > 
> > I remember having problems with OpenSSH's configure back when we used
> > autoconf-2.59 to generate it.
> 
> Can you check whether that configure script has the problematic EOF
> marker crossing a 1024 byte boundary?

As you note below, the _ of the closing the _ASBOX marker is on a 1024
byte boundry.
 
> > > But what is the problem really?  Is it here-doc EOF markers used both in
> > > a shell script as well as in a here-doc in that shell script?  Or is it
> > > here-docs of a certain size, or with certain contents?  Without
> > > understanding that, we cannot know if the workaround is effective or
> > > we'll introduce another similar issue in 2.66.
> > 
> > I've also attached the config.status from autoconf-2.59 test 063 that
> > shows the problem.
> 
> autoconf-2.65/tests/testsuite.dir/226/inner/configure and
> autoconf-2.59/tests/testsuite.dir/063/configure both
> have in common that a 1024-byte boundary lies at the start of, or
> inside, the _ASBOX marker.
> 
> > I don't really know what the problem is. It seems to be if there are
> > here-doc within a here-doc in a shell script of a certain size
> > with certain contents.
> 
> I suspect that here-doc within here-doc is a red herring; but I'm not
> sure.  Can you check whether the problem disappears if, in the 2.65
> test, you change the one problematic EOF marker pair from _ASBOX to a
> different string with the same length?

Changing to a different string of the same length produces the same error.
 
> > I can add or subtract a space and get different results.
> 
> Thanks for these tests.  Indeed sounds very similar to the bug we
> documented in the long paragraph in
>   info Autoconf Here-Documents
> 
> Which ksh version is this?  (Do 'set -o emacs' and then CTRL-V inside
> ksh to find out.)

Version M-12/28/93e-SCO

> 
> * Tim Rice wrote on Fri, Jan 15, 2010 at 06:56:30AM CET:
> > On Tue, 12 Jan 2010, Tim Rice wrote:
> > > 
> > > Unfortunatly we're already selecting the best (least tests fail) shell
> > > that ships with the system.
> > > /bin/sh will pass test 127 (and 226) but fails 123 133 134 and 211.
> > 
> > Now I'm not so sure.
> >  
> > > I started working on building bash-4.1 last night.
> > > When I get that built, I'll see if it works any better.
> > 
> > I re-ran test 127 with
> >    gmake check TESTSUITEFLAGS='-v -d -x 127 CONFIG_SHELL=/bin/sh'
> > and
> >    gmake check TESTSUITEFLAGS='-v -d -x 127 CONFIG_SHELL=/opt/bin/bash'
> > 
> > See attached sh-testsuite.log.gz and bash-4.1-testsuite.log.gz
> > In both cases the inner tests fail but the last line says test 127 is ok.
> > Strange.
> 
> The inner tests 4-6 are supposed to fail.  It is a bug if they pass.

Ah, it faked me out. Thanks for the clarification.

> 
> > At least ksh says 127 failed.
> 
> Well yes, that's the ksh issue with output not ending with a newline.
> 
> Cheers,
> Ralf

-- 
Tim Rice                                Multitalents    (707) 887-1469
address@hidden

bash-4.1-testsuite.log.gz
Description: Binary data

bash3.1.17-testsuite.log.gz
Description: Binary data