master copy of ltcf-c.sh, ltcf-gcj.sh, ltmain.sh?

master copy of ltcf-c.sh, ltcf-gcj.sh, ltmain.sh?

[Michael Elizabeth Chastain]

Hi guys,

I have an issue with some of the top level configuration files:

  ltcf-c.sh
  ltcf-gcj.sh
  ltmain.sh

The issue is that these files use some commands like 'tail -1' which are
not allowed by POSIX 1003.1-2001.  (This is coming from PR gdb/527 and
PR gdb/894 if you are curious).

Who owns the master copy of these files, and where do I go to file a PR
against them?  I am guessing that the answer is 'the gcc copy is the
master copy' and 'file a gcc PR with category=other', but I need to
check this.

Michael C

Re: master copy of ltcf-c.sh, ltcf-gcj.sh, ltmain.sh?

[Andreas Tobler]

Alexandre Oliva wrote:
> On Mar  3, 2003, Andreas Tobler <toa@pop.agri.ch> wrote:
> 
> 
>>In case it'll take a while, few weeks, would it be a try worth to
>>patch the missing parts for darwin in ltcf-c, c++ and gcj?
> 
> 
> Might be, but I'm seeing a lot of activity in the Darwin front in
> libtool-patches, so maybe we'd better wait until it settles down.
> 
Ok, then I play local and hope in the near future we will see something :)

Thanks,

Andreas

Re: master copy of ltcf-c.sh, ltcf-gcj.sh, ltmain.sh?

[Alexandre Oliva]

On Mar  3, 2003, Andreas Tobler <toa@pop.agri.ch> wrote:

> In case it'll take a while, few weeks, would it be a try worth to
> patch the missing parts for darwin in ltcf-c, c++ and gcj?

Might be, but I'm seeing a lot of activity in the Darwin front in
libtool-patches, so maybe we'd better wait until it settles down.

-- 
Alexandre Oliva   Enjoy Guarana', see http://www.ic.unicamp.br/~oliva/
Red Hat GCC Developer                 aoliva@{redhat.com, gcc.gnu.org}
CS PhD student at IC-Unicamp        oliva@{lsd.ic.unicamp.br, gnu.org}
Free Software Evangelist                Professional serial bug killer

Re: master copy of ltcf-c.sh, ltcf-gcj.sh, ltmain.sh?

[Andreas Tobler]

Alexandre Oliva wrote:
> On Mar  3, 2003, "Andreas Tobler" <toa@pop.agri.ch> wrote:
> 
> 
>>But I expect that it isn't that simple since the cvs libtool relies on autoconf
>>2.5x. And we in gcc don't do yet.
> 
> 
> That's exactly the main stumbling block at the moment.  Unfortunately,
> I'm yet to review Klee Dienes' patch to adopt a newer version of
> autoconf, so I can't even tell whether we're just a review away from
> being able to do it or whether it will require major rework.

In case it'll take a while, few weeks, would it be a try worth to patch 
the missing parts for darwin in ltcf-c, c++ and gcj?

Though, I'm not ready yet, since it is not that easy to satisfy all sub 
libraries...

Andreas

Re: master copy of ltcf-c.sh, ltcf-gcj.sh, ltmain.sh?

[Alexandre Oliva]

On Mar  3, 2003, "Andreas Tobler" <toa@pop.agri.ch> wrote:

> But I expect that it isn't that simple since the cvs libtool relies on autoconf
> 2.5x. And we in gcc don't do yet.

That's exactly the main stumbling block at the moment.  Unfortunately,
I'm yet to review Klee Dienes' patch to adopt a newer version of
autoconf, so I can't even tell whether we're just a review away from
being able to do it or whether it will require major rework.

-- 
Alexandre Oliva   Enjoy Guarana', see http://www.ic.unicamp.br/~oliva/
Red Hat GCC Developer                 aoliva@{redhat.com, gcc.gnu.org}
CS PhD student at IC-Unicamp        oliva@{lsd.ic.unicamp.br, gnu.org}
Free Software Evangelist                Professional serial bug killer

Re: master copy of ltcf-c.sh, ltcf-gcj.sh, ltmain.sh?

[Andreas Tobler]

>-- Original Message --
>Date: Mon, 3 Mar 2003 08:53:46 -0600
>From: Michael Elizabeth Chastain <mec@shout.net>
>To: aoliva@redhat.com, bonzini@gnu.org
>Subject: Re: master copy of ltcf-c.sh, ltcf-gcj.sh, ltmain.sh?
>Cc: gcc@gcc.gnu.org
>
>
>Hi Alexandre,
>
>> It is possible that libtool still uses tail -1, and that any fixes to
>> CVS mailing to this effect could be easily merged into our copy, but
it
>> has to be fixed there first.
>
>Paul Eggert has already fixed the 'head -1' and 'tail -1' and
>other non-posix commands in the CVS version of libtool.
>
>What are the prospects of importing a new libtool into gcc?

I'm also interested in this question since the cvs libtool contains lot
of stuff needed for darwin and dylibs.

I hand ported those changes to a 3.3 branch and have now a gcc with shared/dynamic
libraries for darwin. (not completed yet but a good start)

But I expect that it isn't that simple since the cvs libtool relies on autoconf
2.5x. And we in gcc don't do yet.

Andreas

Re: master copy of ltcf-c.sh, ltcf-gcj.sh, ltmain.sh?

[Michael Elizabeth Chastain]

Hi Alexandre,

> It is possible that libtool still uses tail -1, and that any fixes to
> CVS mailing to this effect could be easily merged into our copy, but it
> has to be fixed there first.

Paul Eggert has already fixed the 'head -1' and 'tail -1' and
other non-posix commands in the CVS version of libtool.

What are the prospects of importing a new libtool into gcc?

Michael C

Re: master copy of ltcf-c.sh, ltcf-gcj.sh, ltmain.sh?

[Alexandre Oliva]

On Feb 25, 2003, Paolo Bonzini <paolo.bonzini@polimi.it> wrote:

> They live in libtool's CVS head.

Actually, these files are modified copies of the libtool
multi-language branch, that is now merged into libtool CVS mainline
and heavily modified since then.  It is possible that libtool still
uses tail -1, and that any fixes to CVS mailing to this effect could
be easily merged into our copy, but it has to be fixed there first.

-- 
Alexandre Oliva   Enjoy Guarana', see http://www.ic.unicamp.br/~oliva/
Red Hat GCC Developer                 aoliva@{redhat.com, gcc.gnu.org}
CS PhD student at IC-Unicamp        oliva@{lsd.ic.unicamp.br, gnu.org}
Free Software Evangelist                Professional serial bug killer

Re: master copy of ltcf-c.sh, ltcf-gcj.sh, ltmain.sh?

[Paolo Bonzini]

They live in libtool's CVS head.

-- 
|_  _  _ __
|_)(_)| ) ;'_.