Re: GLIBCPP_CHECK_LINKER_FEATURES libstdc++/acinclude.m4

[Phil Edwards <pedwards at disaster dot jaj dot com>]

On Sun, Jul 08, 2001 at 02:24:08PM -0500, Gordon Sadler wrote:
> The above function appears to be 'turned off' on both head and 3.0
> branch. Using cvs annotate, it appears Ben Kosnik made a change that
> effectively turns this function off, there are some comments around
> libstdc++/acinclude.m4:252 that appears to say this won't work and in
> some cases causes a core on solaris.

Yep.  The garbage collection in GNU ld is too hungry:

    http://gcc.gnu.org/ml/gcc/2001-03/msg00401.html


> In this AC_DEFUN there are two things that drew my attention:
> 1. The use of ac_cv_prog_gnu_ld which is not valid here. Instead
>    libtool.m4 defines/uses lt_cv_prog_gnu_ld.

Hmm.  I think I got that by looking at the config.cache file, but there
have been plenty of changes since.  So it might have changed, sure.

> 2. The test (:242) of $ac_cv_prog_gnu_ld = "notbroken". I can't find any
>    usage of assigning 'notbroken' to any variable anywhere.

There isn't.

> >From some of the comments, it looks like this is all intentional, but..
> The Changelog entries for 5 Dec 00, 11 Oct 00, and 17 Oct 00 do not
> sound like this is the intent. Phil Edwards Changelog from 11 Oct says
> comment out some code in a test, but Ben made a change the same day to
> the same parts.

Nope, it's intentional.  Rather than commenting it out, it's rigged so
that it always fails.

> If this AC_DEFUN really should be active I suggest
> s/ac_cv_prog_gnu_ld/lt_cv_prog_gnu_ld/ and whatever check was being
> performed for 'notbroken' is redone as currently that does not exist.

No, that's the whole point.  The test /can't/ pass.  We don't want it to.

There is a good reason why this isn't "formally" commented out, but I
don't recall offhand what it is.


Phil

-- 
Would I had phrases that are not known, utterances that are strange, in
new language that has not been used, free from repetition, not an utterance
which has grown stale, which men of old have spoken.
                                     - anonymous Egyptian scribe, c.1700 BC