[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: unexpected editing of argument to AH_BOTTOM
From: |
Stepan Kasal |
Subject: |
Re: unexpected editing of argument to AH_BOTTOM |
Date: |
Fri, 13 Oct 2006 12:38:52 +0200 |
User-agent: |
Mutt/1.4.2.1i |
Hello Paul,
On Thu, Oct 12, 2006 at 02:52:05PM -0700, Paul Eggert wrote:
> "John W. Eaton" <address@hidden> writes:
> > I'm wondering whether this behavior is a bug or feature.
Well, in any case, this is a trap which many people fell into.
At least the documentation has to be improved; a proposed patch
attached for a review.
> I'd say it's a bug. Fixing it coherently may not be that easy,
> though.
I'm not sure how the ``fixed'' behaviour should look like.
Imagine you want to ``correct'' the detectes defines, and add the
following at the end of config.h.in:
/* For HP/UX with gcc */
#if defined(hpux) || defined(_HPUX_SOURCE)
#undef HAVE_TZSET
#define HAVE_TZSET 1
#define _TZSET 1
#endif
It is inevitable that if HAVE_TZSET was AC_DEFINEd, then the #undef
get's replaced. So this trap will inevitably be there.
At the end of the proceesing of the template, all remaining #undef's
get commented out. A comment in status.m4 explains that:
# Then replace #undef with comments. This is necessary, for example,
# in the case of _POSIX_SOURCE, which is predefined and required
# on some systems where configure will not decide to define it.
So at least _some_ #undef's have to be commented out.
We could perhaps comment out only the #undef's corresponding to the
AC_DEFINEs in configure sources. (But that would not help with the
above example.)
Have a nice day,
Stepan
autoconf-20061013-ah-doc.patch
Description: Text document